Category: Startup Funding News

Thea Energy is a New Jersey fusion company building stellarator reactors with software-controlled flat magnets instead of the twisted coil systems that have made stellarators brutally hard to manufacture. The startup has now raised an oversubscribed $100 million Series B led by Thomas Tull’s U.S. Innovative Technology Fund, a round that pushes it into the better-funded tier of private fusion companies. That matters because fusion doesn’t usually fail on ambition. It fails when exotic physics runs into impossible hardware, cost, and maintenance demands. Thea was founded in 2022 by Brian Berzin, David Gates, and Matt Miller, and the company’s pitch is that smarter magnet architecture might finally make stellarators practical enough to leave the lab.

What does Thea Energy actually build?

Thea Energy builds a planar-coil stellarator. In plain English, it uses arrays of high-temperature superconducting flat magnets and software controls to create the 3D magnetic fields needed to confine plasma. It does that without relying on the famously awkward custom-shaped modular coils that define older stellarator designs. That shift moves complexity out of hardware and into software. A big deal if you’re trying to build lots of identical parts instead of one-off precision sculptures.

Its first major machine is Eos, which Thea describes as a “power plant relevant” integrated demonstration system. Eos is designed as a deuterium-deuterium neutron source that can run in steady state and produce isotopes, including tritium and medical radioisotopes. It also gives the company a way to prove its core architecture on something much closer to a commercial plant than a science experiment. Helios is the follow-on commercial power plant.

What stands out here isn’t just the magnets. It’s the control layer. Thea says the system can optimize operating points in real time and correct for changing conditions. It also keeps the machine adaptable instead of locked into one fixed geometry. That’s a sharp contrast with classic stellarators, where a lot of the performance is baked into hard-to-change 3D coil shapes from day 1.

Maintenance is part of the pitch too. Thea’s geometry is supposed to allow sector-based access, so operators can remove large sections for service with less downtime than older stellarator concepts. One small but important wrinkle: early Thea designs talked about 12 encircling magnets, but later versions dropped that feature. The current story is less about a fixed outer ring and more about programmable planar field control.

Who founded Thea Energy and why does it matter?

The founding story

Thea Energy came out of Princeton Plasma Physics Laboratory and Princeton University work on magnet-array-based stellarator designs. David Gates developed the underlying stellarator magnet array technology at PPPL through the ARPA-E BETHE program, and the company spun out in 2022 to commercialize that research. That origin matters because Thea isn’t trying to invent an entirely new confinement concept from scratch. It’s taking a known fusion path and trying to make it buildable.

Why this founding team has real market fit

Brian Berzin brings the commercialization angle. He previously served as VP of Strategy at General Fusion and also has experience in venture capital, growth equity, private equity, and electrical engineering startups. That’s useful because fusion startups don’t just need plasma talent. They need people who understand fundraising and industrial partnerships. They also need people who know how hardware companies die when capital planning goes sideways.

Gates brings the scientific credibility. He has 30+ years in fusion research across stellarators and tokamaks, served as head of advanced projects at PPPL, held a senior research role at Princeton, and had produced 200+ publications with 7,500+ citations by 2025. If Berzin is the bridge to the market, Gates is the reason investors can take the engineering thesis seriously.

Traction, fundraising, and the current roadmap

Thea announced its latest raise on May 27, 2026. USIT led the $100 million Series B, with participation from General Innovation Capital Partners, Linse Capital, Calm Ventures, Climate Capital, Divergent Capital, Emerald Technology Ventures, Gaingels, Idemitsu Kosan, Overlay Capital, Timescale Ventures, and What If Ventures. The company had already raised a $20 million Series A, and the new round brings total private investment to $130 million, according to the source article.

The money is earmarked for expanding magnet manufacturing and starting construction of Eos next year. By early 2026, Thea had also won DOE certification for its Helios preconceptual design, said it was talking with 5 states about siting Eos, and described itself as an 80+ person team of engineers, scientists, and commercialization staff. The public target is ambitious: complete Eos in 2030, then bring the Helios commercial plant online in 2034.

How Thea Energy compares with other fusion startups

The obvious benchmark is Commonwealth Fusion Systems. CFS is following the tokamak route and plans to use new capital to finish SPARC. It also wants to advance its first ARC plant in Virginia, and now says it has raised $3 billion in total. That’s a completely different scale from Thea. It shows how hard it is for any new fusion company to stay relevant without a credible manufacturing shortcut.

Type One Energy is a closer conceptual comparison because it’s also pursuing a stellarator. In January 2026, TechCrunch reported that Type One had raised an $87 million convertible note, bringing its total venture backing to more than $160 million, though its business model leans more toward selling core technology to utilities and power providers. Thea’s distinction is the planar-coil, software-heavy architecture. Simpler parts. More configurability. A maintenance story that’s easier to explain to plant operators.

There’s also the legacy comparison. Traditional stellarator alternatives lean on extremely complex 3D magnetic coil sets that are difficult to build, align, and service. Thea is betting investors will back software-defined field shaping and repeatable magnet manufacturing over bespoke machine craftsmanship.

Why Thea Energy’s $100M round matters

This round matters because it shifts Thea from “interesting reactor concept” toward “actual industrial build program.” Expanding magnet manufacturing is a very different milestone from publishing papers or running bench tests. It means the company now has to prove that its simplification story survives contact with real production tolerances, real supply chains, and real schedules.

Eos is also a smart midpoint if Thea can execute. A neutron-source system that can run in steady state and produce useful isotopes gives the company something closer to an interim commercial path while de-risking the Helios architecture. That’s a lot more believable than promising grid power first and figuring out revenue later.

But let’s be honest: $100 million is big for a young fusion startup, not big for fusion full stop. The round buys Thea time and hardware progress. It also buys talent. It doesn’t buy certainty. The company still has to show that its programmable magnet thesis works at scale, not just in prototypes and design packages.

Is fusion finally becoming a real market?

Fusion still isn’t a market in the conventional sense. It’s a race to make one. But the structure around it is getting more real. The U.S. Department of Energy’s current fusion roadmap is explicitly aimed at accelerating commercialization by the mid-2030s and scaling the domestic private fusion sector in the 2030s.

Money has followed that shift. In the Fusion Industry Association’s 2023 industry report, private fusion companies had attracted more than $6 billion in investment, up $1.4 billion year over year, and the U.S. alone counted 25 private fusion companies. That doesn’t mean fusion is solved. It does mean investors, governments, and power-hungry industries are treating it less like science fiction and more like a manufacturing challenge with a clock on it.

That timing helps Thea. Not because fusion got easier overnight, but because grid demand is changing what capital wants. Baseload, carbon-free power is suddenly a very hot word again. A startup that can argue for lower capital cost and simpler maintenance will get a longer hearing now than it would have 5 years ago. Software-tunable hardware helps too.

Thea Energy outlook

Thea Energy now has real money, a differentiated stellarator thesis, and a clearer path to its first serious machine. That’s enough to make it one of the more credible second-tier fusion bets in the U.S., though still far behind the capital base of the category’s biggest names. The next thing to watch isn’t another concept rendering. It’s whether Eos construction actually starts in 2027 and whether Thea can show that its magnet architecture scales from smart idea to reliable hardware.

Read how Triomics raised $22M in Series B funding to use oncology-focused AI for clinical trial matching, chart prep, and cancer registry workflows.

FAQ

– What was Thea Energy’s latest funding round?  

 Thea Energy raised an oversubscribed $100 million Series B on May 27, 2026. USIT led the round, and the cash is meant to expand magnet manufacturing and push Eos into construction.

– How does Thea Energy’s fusion reactor work?  

 Thea is building a stellarator that uses software-controlled planar HTS magnets to shape the magnetic field around plasma. The idea is to replace much of the hard-to-manufacture 3D coil complexity of older stellarators with flatter, more repeatable components. A control stack can tune the field in real time.

– How did Thea Energy’s founders get into fusion?  

 The company’s leadership blends lab science with fusion commercialization. Brian Berzin came from General Fusion and finance-backed startup work, while David Gates spent decades in fusion research at PPPL and Princeton before spinning the core technology into Thea in 2022.

– Is Thea Energy a tokamak or a stellarator company?  

 Thea Energy is a stellarator company, not a tokamak company. That matters because stellarators are known for steady-state stability, but they’ve historically been dragged down by magnet complexity. That’s the problem Thea is trying to solve with its planar-coil architecture.

Triomics builds software that uses oncology-specific AI to read messy cancer records and automate work like clinical trial matching, chart prep, and registry reporting. The oncology AI startup has now raised $22 million in Series B funding. Battery Ventures led the round, with returning backers Nexus Venture Partners, Lightspeed, Y Combinator, and others joining in. Founded in 2021 by CEO Sarim Khan and CTO Hrituraj Singh, the company is chasing a very real bottleneck: cancer patients are living longer, which is great news, but it also means staff are stuck digging through years of physician notes, pathology reports, imaging, and even scanned faxes before they can do basic operational work. Khan put it plainly in the source interview: “We have seen medical records [with] thousands of pages of information.”

What does Triomics’ oncology AI platform actually do?

At the center of the product is OncoLLM, Triomics’ oncology-focused AI framework. It isn’t a single model. It’s a system of 8 models, ranging from 3B to 72B parameters, designed to reason at the patient level rather than just summarize one document at a time. That matters in oncology, where the signal is spread across a long timeline and buried in different record types.

The clearest example is PRISM, Triomics’ trial-enrollment software. In practice, PRISM pulls from both structured and unstructured EHR data. It checks a patient’s chart against active trial criteria and generates match summaries that coordinators and physicians can review before the visit. In one deployment at the Medical College of Wisconsin Cancer Center, the system screened 100% of upcoming visits across 5 disease teams against more than 100 recruiting trials after just a 2-hour onboarding workflow for coordinators.

That’s only one layer of the product now. Triomics started with clinical trial matching, but as large language models got better, it expanded into verifiable patient summaries for appointment prep. It also added data-curation tools that support quality reporting, cohort analysis, precision oncology, and cancer registry workflows. The company’s 2024 materials also described Harmony, a product for curation and reporting, alongside software embedded into health-system EHRs for task-specific work rather than generic chatbot-style assistance.

The before-and-after is pretty simple. Before Triomics, staff manually read charts for hours. They extract evidence by hand and repeat the same review process across every trial, visit, or reporting obligation. After Triomics, the software does the first pass in minutes and surfaces evidence-backed summaries inside the tools clinicians already use. In published results highlighted by the company, PRISM achieved more than 95% accuracy in placing the correct trial within its top 3 recommendations.

Who founded Triomics, and why does this oncology AI startup have investor backing?

The founding story

Triomics was founded in 2021 by Sarim Khan and Hrituraj Singh. The pairing makes sense on paper and, honestly, that’s rare. Khan came from chemical engineering and research experience in tissue engineering and neuroscience, while Singh had worked at Adobe Research on language models and reinforcement learning. They were also college friends, which usually helps when you’re trying to build a company in a category where the sales cycles are long and the consequences of errors are brutal.

Their original wedge was clinical trial matching. And that wasn’t random. They saw that a lot of hospital software could already work with the tidy 20% of health data stored in structured fields, but the harder 80% sat in free text and document chaos. Oncology was an extreme version of that problem. So Triomics went after the specialty first and then widened the product as LLMs became capable enough to handle more than trial screening.

Why the founders fit this category

Khan brings the biomedical lens. Singh brings the model-building depth. That’s a useful mix when your customer doesn’t want a flashy copilot — they want software that can survive review by oncologists, trial coordinators, and compliance teams.

Triomics also didn’t build in isolation. The company worked with Medical College of Wisconsin researchers on OncoLLM. It has also leaned hard into consortium-style validation around safety and benchmarking. Its leadership has pointed to COLT, a collaboration involving more than 20 NCI-designated cancer centers and Ci4CC, as part of that effort. In a field like oncology, that kind of institutional co-development matters a lot more than a polished demo.

Traction and the funding

Triomics’ enterprise customer base expanded 4x over the past year, helping push annualized recurring revenue up 10x. The customer list is getting serious. Memorial Sloan Kettering and Yale Cancer Center already use Triomics, while Y Combinator says the company is trusted by 4 of the top 10 Best Hospitals for Cancer ranked by U.S. News. Mount Sinai also rolled out PRISM systemwide in early 2026, becoming the first NCI-designated Comprehensive Cancer Center in New York City to deploy the tool for enterprise clinical trial matching.

The new money follows Triomics’ $15 million Series A in May 2024. That earlier round included Lightspeed, Nexus Venture Partners, General Catalyst, and Y Combinator. Now Battery Ventures is stepping in to lead the Series B. Investors see something bigger here than a single workflow feature.

How does Triomics compare with Abridge, Nuance, and legacy workflows?

Triomics’ closest overlap with Abridge and Microsoft’s Nuance products is chart summarization, but the products aren’t really the same thing. Abridge turns clinical conversations into documentation and billable outputs. Microsoft Dragon Copilot combines dictation, ambient listening, and workflow support across general clinical documentation. Triomics is aiming at a narrower but thornier problem: making sense of longitudinal oncology records for tasks like trial eligibility, pre-charting, quality programs, and tumor registry submission.

The real incumbent still isn’t software. It’s manual labor. Nurses, coordinators, and admin staff are still reviewing huge charts by hand because generic AI tools don’t reliably understand oncology nuance. That’s the company’s core pitch, and it’s why Khan argues cancer centers such as MSK and Yale have chosen Triomics over broader-purpose assistants. The bet is that domain specificity beats generality when the workflow is specialized, highly regulated, and expensive to get wrong.

Why this oncology AI startup’s $22M round matters

The obvious read is that Triomics now has the capital to keep broadening from trial matching into a fuller oncology operations stack. That’s where the company has been heading for a while — first matching, then patient summaries, then registry and quality workflows. If that works, Triomics stops being a point tool and starts becoming infrastructure for cancer centers.

That matters for customers because oncology admin work doesn’t sit neatly in one place. A patient’s case can stretch over years, bounce across sites of care, and generate a chart that’s too dense for generic summarizers to handle well. A tool that can shrink appointment prep time and automate mandatory reporting buys back something rare in cancer care: staff attention. That’s valuable even before you get to trial enrollment.

There’s also a harsher reason this round matters. Healthcare AI is crowded now. Ambient scribes are everywhere. EHR vendors are shipping their own copilots. So for a startup to raise a Series B here, it usually needs more than a cool model — it needs evidence, workflow fit, and customers that will stick. Triomics understands that. Its emphasis on oncology-specific training, published validation work, and deployment inside real cancer centers suggests a company trying to build defensibility the boring way.

How big is the market for oncology AI software?

The most relevant adjacent market here is oncology information systems. That market was worth about $2.94 billion in 2024 and is projected to reach roughly $4.69 billion by 2030, with an 8.1% CAGR from 2025 through 2030. North America accounted for 39.7% of the market in 2024, which helps explain why startups like Triomics are targeting U.S. cancer centers first.

The timing also makes sense at the workflow level. Fewer than 10% of adult oncology patients enroll in clinical trials, and one reason is painfully simple: matching is still labor-intensive, fragmented, and easy to miss in day-to-day care. When records include free-text notes, pathology, imaging, genomic reports, and faxed documents, software that can actually reason across all of it has a much bigger opening than another generic AI note tool.

The broader healthcare AI market is sending a mixed signal. On one hand, ambient scribes became a $600 million category in 2025 after 2.4x year-over-year growth. On the other, Menlo Ventures says switching costs are low, pricing pressure is rising, and customers increasingly expect scribes to expand beyond documentation into more durable workflows. That’s relevant for Triomics because it suggests vertical, workflow-deep products may have a better shot at lasting value than standalone summarization tools.

Conclusion

Triomics looks like the kind of oncology AI startup investors want right now: focused, evidence-heavy, and pointed at work that hospitals already pay people to do manually.

But the next phase is harder than the last one. It’s one thing to prove an oncology model can read a chart. It’s another to become the default workflow layer across trial matching, pre-charting, and registry operations at major cancer centers. The company just raised $22 million to make that case.

Read how Tiea Connectors raised ₹77 Cr in Series A funding to build high-performance electrical connectors and interconnect systems for EV, aerospace, defence, and avionics manufacturers in India.

FAQ

– What funding did Triomics raise in 2026? 

 Triomics raised a $22 million Series B round announced on May 27, 2026. Battery Ventures led the financing, and returning investors included Nexus Venture Partners, Lightspeed, and Y Combinator.

– How does Triomics work for cancer centers? 

 Triomics uses an oncology-specific AI framework called OncoLLM to process both structured and unstructured patient records. It then powers workflow tools such as PRISM for clinical trial matching. Instead of only summarizing a note, it evaluates longitudinal charts, checks eligibility against trial criteria, and produces evidence-backed summaries for coordinators and physicians inside existing clinical workflows.

– Who founded Triomics? 

 Triomics was founded in 2021 by Sarim Khan and Hrituraj Singh. Khan’s background spans chemical engineering plus tissue engineering and neuroscience research, while Singh previously worked at Adobe Research on language models and reinforcement learning before becoming Triomics’ CTO.

– Is Triomics an ambient scribe company or an oncology software company? 

 It’s closer to an oncology software company than a pure ambient scribe vendor. Abridge and Microsoft Dragon Copilot mainly focus on turning clinician-patient conversations into documentation, while Triomics is built around oncology-specific chart review, trial enrollment, pre-charting, quality workflows, and registry reporting inside a market adjacent to oncology information systems.

Tiea Connectors makes high-performance electrical connectors and contact systems for sectors where failure isn’t an option. The Bengaluru-linked hardware startup has raised ₹77 crore in Series A funding led by IvyCap Ventures, with Jamwant Ventures, 8X Ventures, and a set of HNI angel investors also joining the round. That matters because India still imports a lot of precision interconnect hardware for EVs, aerospace, defence, and avionics. Local manufacturers that can design, tool, test, and scale these parts are still rare. Founded in 2020 by Ajith Sasidharan and Punit Shridhar Joshi, Tiea is trying to build that missing layer.

What does Tiea Connectors make and how does it work?

Tiea Connectors designs and manufactures electrical connectors, precision contacts, connector housings, stamped terminals, and cable-and-harness assemblies for OEMs and product teams. In plain English, it builds the small but mission-critical parts that let power and data move reliably inside vehicles, charging systems, drones, industrial electronics, defence equipment, and avionics hardware.

Its workflow is more vertically integrated than what a lot of small hardware suppliers can manage. An OEM can come in with an application requirement. Tiea then handles connector architecture and material selection. It also takes on rapid prototyping, tooling, validation, and scaled manufacturing. That matters. It cuts out the usual mess of splitting work between a design consultant, a toolroom, and a contract manufacturer.

The product mix is unusually broad for a young interconnect company. Tiea lists standard and custom connectors, presstac contacts for EV charging and battery management systems, high-precision machined and stamped parts, gold-plated terminals for harsh environments, in-house injection-moulded connector housings, and custom cable assemblies. Some of its EV-focused contacts are built for high current loads and up to 100,000 mating cycles. That gives you a sense of the reliability bar it’s chasing.

The before-and-after story for customers is obvious. Before, an Indian OEM in a niche category often had to import connector systems, wait on long qualification cycles, and accept whatever catalog part a global supplier was willing to ship. After, the same buyer gets a local engineering partner that can customize around space limits, vibration, heat, compliance needs, and cost targets without sending the whole program overseas.

Who founded Tiea Connectors and what has it built so far?

Founded by two engineers who knew the supply gap

Tiea was founded in 2020 by Ajith Sasidharan and Punit Shridhar Joshi. The two had worked together earlier at HPCL, then left to start the company after seeing how dependent Indian manufacturers still were on imported connector and contact solutions. That origin story fits the product. This isn’t a team that stumbled into hardware because it looked trendy.

The company started by focusing on miniature and customized electronic connectors. It has since widened into high-performance interconnect systems for electric mobility, aerospace, defence, avionics, and other emerging applications. Tiea also works as an original design manufacturer for tech-focused OEMs that want a partner to own connector design and manufacturing, not just assemble to print.

Tiea was incubated at IISc Bengaluru’s Foundation for Science and Innovation Development. It has also been linked with defence-focused development through the iDEX Challenge. That’s a decent signal that its ambitions aren’t limited to commodity parts.

Why the founders had a real shot at this

Ajith Sasidharan is a College of Engineering Trivandrum alumnus and has built his career around operations, business development, and interconnect products. Punit Shridhar Joshi, who is identified publicly as Tiea’s CTO, studied at NITK and brings the technical depth the category demands. That split works well for a manufacturing startup. One founder stays close to execution and growth. The other stays close to engineering and product.

That matters more in connectors than most people think. This category punishes weak process control. You don’t win because your pitch deck sounds smart. You win because tolerances stay tight, materials behave the way they should, tooling doesn’t drift, and a part still works after vibration, thermal cycling, and repeated mating. Tiea’s pitch is basically that it can do the hard bit in-house.

Its internal strengths line up with that claim: product architecture, rapid prototyping, tooling development, material engineering, miniaturisation, precision manufacturing, and application-specific customization. That’s a lot of capability for a startup that’s only been around since 2020.

Early traction, prior rounds, and the Series A

Tiea isn’t operating like a lab project. It is delivering 5 million parts a month, has 70+ active projects, and serves 50+ customers. A prior update around its earlier fundraise said the business had grown 4x in FY25 and was already running at 90% of manufacturing capacity. For a hardware company, that’s the kind of signal investors care about.

It also has a manufacturing facility in Dharwad, Karnataka, with in-house tooling, stamping, moulding, assembly, and testing. That setup is central to the thesis. If you’re selling into defence electronics, EV systems, or aerospace programs, you can’t fake process ownership for long.

This Series A comes after smaller earlier rounds, including a ₹3 crore angel raise in 2022 and a ₹22 crore round announced later as the business scaled. IvyCap Ventures led the new ₹77 crore round. Jamwant Ventures and 8X Ventures also came in, along with select high-net-worth angel investors. Tiea says the money will go into manufacturing expansion and stronger R&D and product engineering. It also plans more automation and technology integration, along with broader scaling in India and overseas.

How Tiea competes in a connector market ruled by imports

Tiea is entering a category long dominated by global connector majors and imported parts catalogs. In practice, Indian OEMs often buy from large multinational suppliers or rely on fragmented local vendors that can machine or mould a part but can’t own the full engineering cycle. That gap is where Tiea is trying to sit.

Its edge isn’t that it’s inventing connectors from scratch. It offers local design ownership and faster tooling iteration. It also handles customization, validation, and scaled manufacturing under one roof. For customers in EVs, aerospace, and defence, that can mean shorter lead times, less import dependence, easier localization, and products built around Indian operating conditions instead of generic global specs.

That’s also why IvyCap’s bet makes sense. Vikram Gupta, the firm’s founder and managing partner, pointed to Tiea’s strength in innovation, precision engineering, product customization, and scalable manufacturing. The investors aren’t just backing a component seller. They’re backing a domestic supplier that could become hard to replace once it’s qualified into critical programs.

Why does Tiea Connectors funding matter now?

Because this round changes the kind of company Tiea can become.

A small connector maker can survive on engineering skill and founder hustle for only so long. Once orders scale, the real pressure moves to automation, repeatability, quality systems, and production throughput. That’s exactly where Tiea says the fresh capital is going. Not into vague brand building. Into machines, process depth, product engineering, and more factory muscle.

There’s a timing angle here too. Ajith Sasidharan has framed the demand pull around “electrification, intelligent systems, and high-reliability applications.” He’s right. Those categories don’t just need more components. They need better interconnects. A drone, an EV platform, or an avionics unit is only as dependable as the weakest connector buried inside it.

The strategic value is bigger than the part itself. If Tiea can move from a promising domestic supplier to a deeply embedded one, it could become part of the supply chain logic for Indian OEMs that want less exposure to imported precision hardware.

How big is the market for high-performance connectors in India?

It’s not a tiny niche anymore. India’s automotive connectors market was valued at $746 million in 2024 and is projected to reach about $1.2 billion by 2033. That’s just one slice of the opportunity, and Tiea isn’t limited to automotive. It also sells into electric mobility, defence, aerospace, drones, consumer electronics, and industrial systems.

The structural trend is simple. Every time a machine gets more electrified, more software-heavy, or more safety-critical, connector content goes up. EVs need high-voltage power connectors and battery-management interconnects. Aerospace and defence systems need ruggedized, high-reliability parts that survive heat, shock, and vibration. Avionics and drones need lightweight connectors that still hold signal integrity. None of that is easy manufacturing.

India’s policy push toward domestic electronics production helps too, but policy alone doesn’t build a supplier base. You still need companies that can do toolmaking and moulding. You also need stamping, validation, and quality control at production scale. That’s why this category is getting more interesting now than it was 5 years ago.

What should you watch after Tiea Connectors funding?

The next test for Tiea won’t be fundraising. It’ll be execution.

Can it turn more factory capacity and automation into consistent quality across EV, aerospace, and defence-grade programs? Can it win deeper design-in positions with OEMs instead of staying a useful but replaceable vendor?

Read how SOND raised $7M to launch Dreambuds, AI sleep earbuds that monitor biometric signals and adjust audio in real time to help users fall asleep and stay asleep.

FAQ

– What is the Tiea Connectors funding round about? 

 Tiea Connectors has raised ₹77 crore in a Series A round led by IvyCap Ventures. Jamwant Ventures, 8X Ventures, and a group of HNI angel investors also participated, and the money is meant to expand manufacturing, strengthen R&D, and add more automation.

– How does Tiea Connectors actually make money? 

 Tiea sells connectors, contacts, precision components, and integrated interconnect solutions to OEMs and manufacturers. It also works as an ODM partner, which means it can take responsibility for product definition and design. It also handles validation, tooling, and scaled production instead of just supplying a standard part.

– Who founded Tiea Connectors? 

 Tiea was founded in 2020 by Ajith Sasidharan and Punit Shridhar Joshi. The founders were former colleagues at HPCL, and the company later grew through IISc incubation and deeper work in EV, defence, and aerospace-grade interconnect products.

– Is Tiea Connectors an EV startup or a defence startup? 

 It’s really an interconnect manufacturing startup that sells into both. EV charging, battery management systems, aerospace, defence, avionics, drones, and industrial electronics all need reliable connector systems, and Tiea is building for that broader category rather than one single end market.

SOND makes AI sleep earbuds that monitor your body overnight and change what you hear in real time. On Wednesday, May 27, 2026, the Boston startup came out of stealth with $7 million in funding and a new product called Dreambuds. The pitch is simple: most sleep gadgets tell you what went wrong after the fact, while SOND wants to intervene while you’re still trying to fall asleep or get back to sleep. The company was founded in February 2022 by CEO Yadid Ayzenberg and CTO Amir Lazarovich, two MIT-connected founders who think sleep audio should act more like a live system than a passive player.

What are SOND’s AI sleep earbuds and how do they work?

Dreambuds are a closed-loop, in-ear sleep system. In practice, that means the earbuds collect 12 physiological signals — including respiration, heart-rate variability, cardiorespiratory coupling, sleep stage, body position, snoring, and seismocardiography — then send that data to a cloud-based AI sleep coach that picks or generates audio in response. It’s not just playing white noise all night. It’s meant to notice what’s happening in your body and adjust the intervention while you sleep.

The user flow is more ambitious than standard sleep earbuds. You take the buds out, they resume your sleep plan. The system can switch programs depending on whether you’re winding down, waking up, or stuck in the middle-of-the-night half-awake zone. SOND says users can talk to the coach with a double tap, ask for sleep insights, or request a specific sleep story. They can also choose a soundscape, breathing exercise, binaural beat session, or another program from its library of 500-plus audio tracks. Podcasts can stream through the case too, if that’s your thing.

There’s another important design choice here: Dreambuds don’t need a phone by the pillow. The charging case includes Wi-Fi and Bluetooth. It also has an OLED display, physical buttons, built-in storage, and a speaker that can still fire an alarm if you fall asleep before putting the earbuds in. That phone-free setup is one of the sharper parts of the pitch, and Ayzenberg’s line about it is memorable: giving an insomniac a phone is “like running an AA meeting in a liquor store.”

SOND has also packed in a few details the launch article only hinted at. The free Core plan includes masking sounds and biometric tracking. It also includes nightly reports. A higher-tier Concierge plan adds personalized coaching, the 500-plus track library, and dream journaling. The buds can work offline with downloaded tracks stored in the case, the app includes a Bluetooth-based Find My Dreambuds tool, and battery life is targeted at up to 9 hours of overnight use. It’s a lot.

Who founded SOND and why now?

The founding story

SOND started in Boston in February 2022, but the founders’ connection goes back much further. Ayzenberg and Lazarovich met at MIT about 14 years earlier in a way that sounds almost too on-brand: Lazarovich had just moved into a family dorm without a mattress, and Ayzenberg gave him one from his room. That random sleep-related favor turned into a long friendship, then a company built around sleep.

Ayzenberg’s case for starting SOND came out of his time at Bose. He had led Bose’s sleep products business, launched Sleepbuds II, and spent years hearing a similar request from customers: they didn’t just want masking audio, they wanted sensing, coaching, and actual help improving sleep. Back then, he says, the hardware wasn’t ready to squeeze that many sensors into a tiny earbud without wrecking battery life. By the time Bose stepped away from the category, the technical constraints had shifted enough to make a new attempt possible.

Why these founders fit this category

Ayzenberg looks like a category-native founder because, frankly, he is one. Before Bose, he founded The Sync Project, a Boston startup that mapped music to physiological signals such as heart rate and heart-rate variability. Bose acquired that company in February 2018, which pulled him deeper into the overlap between audio, biosignals, and sleep. He also spent time at the MIT Media Lab, where his work sat close to affective computing and physiological sensing.

Lazarovich brings the systems side. The source article identifies him as a former senior software engineering manager at Google, and SOND frames the company as built by engineers from Bose, Google, and MIT. That matters because Dreambuds isn’t just an audio gadget. It’s earbuds and sensors. It’s embedded hardware, a connected case, cloud processing, voice interaction, and software that has to work when the user is half asleep and annoyed. Brutal stack.

Funding, traction, and where SOND sits against rivals

E14 Fund, Crosslink Capital, Ubiquity Ventures, Alumni Ventures, Meach Cove Capital, and Boston Scientific co-founder John Abele backed the financing. SOND hasn’t disclosed much on commercial traction yet because Dreambuds are still prelaunch. It has said it has run comfort studies and betas, is taking reservations now, plans a crowdfunding campaign, and is aiming for mass production in Q2 2026, with customer availability pegged to mid-2026 on the company’s FAQ.

The obvious direct rival is Ozlo, another sleep-earbud company created by ex-Bose engineers. Ozlo’s current product leans hard into passive blocking and streaming. It also includes biometric sensing, a smart case, and phone-free playback modes. The company announced a $12 million round in October 2024 on top of roughly $8 million raised through crowdfunding. Then there’s Soundcore, whose Sleep A20 sells for $179.99 and emphasizes passive noise blocking, side-sleeper comfort, sleep analytics, app control, and up to 14 hours in sleep mode.

That’s where SOND’s positioning gets interesting. Ozlo and Soundcore mostly help you block noise and stream audio. They also let you review sleep data. SOND is trying to sell something more aggressive: not sleep earbuds as a comfort accessory, but AI sleep earbuds that sense, decide, and intervene. Ayzenberg’s own framing is that Dreambuds are not what a hypothetical Bose Sleepbuds III would have been.

Why does this AI sleep earbuds round matter for SOND?

Because hardware is expensive, and sleep hardware with custom sensing is even worse.

A lot of startup funding headlines blur together. This one doesn’t. Dreambuds combine miniaturized sensors and audio hardware. They also rely on voice interaction, a networked charging case, cloud AI, and a companion app. Getting that from prototype to reliable consumer hardware is a capital problem as much as a product problem. $7 million won’t buy infinite runway, but it does buy time to turn a clever demo into a manufacturable device.

It also gives SOND a shot at building a business that isn’t pure hardware margin. The Core and Concierge split suggests the company wants recurring software revenue layered on top of the earbuds themselves. Investors tend to like that structure for good reason. If the product works, SOND isn’t limited to selling a pair of buds once and hoping customers come back in 3 years.

There’s a subtler signal here too. Bose exited sleep wearables. Ayzenberg still came back to the category anyway. That suggests he thinks the earlier failure was about timing and product scope, not a dead market.

How big is the market for AI sleep earbuds?

The raw market numbers are big enough to attract a lot of builders. Grand View Research sizes the global sleep aids market at $49.1 billion in 2025 and projects it to reach $95.2 billion by 2033, with North America holding a 41.4% share in 2025. That doesn’t mean sleep earbuds alone are a $49 billion business. But it does show why investors keep circling anything that sits between consumer audio, wellness, and sleep improvement.

The demand problem is also very real in the U.S. A CDC data brief published in April 2026 found that 30.5% of American adults got less than 7 hours of sleep in 2024. The same report said 15.4% had trouble falling asleep most days or every day, while 18.1% had trouble staying asleep. That’s a huge addressable group.

Timing matters too. Consumers are already used to wearables that score their readiness, chart their sleep stages, or tell them they had a rough night. What’s changing is the hardware and software stack: sensors are smaller, earbud form factors are more accepted, and cloud-connected systems can react in the moment instead of waiting until morning. SOND didn’t invent that shift, but it’s trying to push it one step further — from sleep tracking to sleep intervention.

Will AI sleep earbuds become more than a niche?

SOND has a credible founder story, a product idea that’s actually distinct, and enough capital to prove whether Dreambuds are more than a smart pitch deck.

But this category is unforgiving. People will tolerate buggy social apps. They won’t tolerate earbuds that die at 4 a.m., fall out, or overcomplicate bedtime. If SOND can ship reliable AI sleep earbuds on the timeline it’s promised, it could help redefine what sleep wearables are supposed to do.

Read how WeRoad raised a $58M Series C led by Airbnb to turn group travel and local meetups into a social platform for young travelers looking to connect beyond traditional booking sites.

FAQ

– What funding did SOND raise for Dreambuds? 

 SOND raised $7 million when it emerged from stealth on May 27, 2026. The investors included E14 Fund, Crosslink Capital, Ubiquity Ventures, Alumni Ventures, Meach Cove Capital, and Boston Scientific co-founder John Abele, which gives the round a mix of MIT ties, consumer-tech backing, and medtech credibility.

– How do SOND Dreambuds work? 

 Dreambuds are AI sleep earbuds that collect 12 physiological signals while you sleep and feed them into a cloud-based coach that changes the audio program in real time. The system can deliver masking sounds and guided exercises. It can also serve sleep stories and other audio responses, and it’s built to run from a connected charging case so users don’t need to keep grabbing their phone at night.

– Who founded SOND? 

 SOND was founded in February 2022 by Yadid Ayzenberg and Amir Lazarovich. Ayzenberg previously founded The Sync Project, which Bose acquired in 2018, and later led Bose’s sleep products work; Lazarovich came from Google and brings the software and systems background this product stack needs.

– Is SOND a sleep tech company or an audio hardware startup? 

 It’s really both, but sleep tech is the better label. Dreambuds sit inside the broader sleep aids market, which Grand View Research values at $49.1 billion in 2025, yet the company’s claim is that it’s building intervention-focused sleep wearables rather than just another pair of audio accessories.

WeRoad is a Milan-based social travel company that sells age-banded group trips and local meetups for people who want to travel with strangers and actually click with them. Its WeRoad Series C round brings in $58 million, led by Airbnb, to fund a U.S. expansion that starts in Austin. The company is chasing a real problem: once people leave college and start working, finding the right travel companions gets weirdly hard. Founded in 2017 by Paolo De Nadai, Fabio Bin, and Erika De Santi, WeRoad is betting that a travel brand can behave more like an offline social network than a booking site.

What is WeRoad and how does it work?

WeRoad basically packages group travel as a social product. Users pick trips by destination, vibe, activity level, length, and budget, then join a small group made up of people in a similar age band. The company’s positioning is blunt: the group leader handles the plan, and the traveler gets to live the trip.

The customer flow is more specific than a normal tour operator pitch. Before booking, travelers can see how many people have already joined a departure and, after logging in, preview basic group details like ages and gender mix. Once booked, they get a coordinator who manages transport and timing. That person also handles accommodation, restaurant bookings, and meeting points. The coordinator opens a WhatsApp group about 2 weeks before departure so the group can start talking before anyone gets on a plane.

That structure removes a lot of the annoying manual work that usually kills group travel—planning, herding, booking, and awkward first-contact logistics. WeRoad also standardizes a bunch of trip elements, including accommodation and internal flights on some itineraries. Insurance and parts of the activity schedule are included too, while international flights stay separate so travelers keep some flexibility. Trips usually run with 8 to 15 people, and the classic itinerary sits around 10 to 12 days. Shorter weekend formats are now part of the funnel.

Then there’s WeMeet, which widens the product from trips to local social life. The app recommends events based on a user’s city and interests. It lets people connect with other attendees and makes it easy to confirm attendance and manage participation. That matters because WeRoad isn’t trying to sell one-off holidays anymore—it’s trying to build a recurring community layer that starts at home and can convert into travel later.

Who founded WeRoad and why did they start it?

The founding story

The origin story is unusually personal for a travel startup. De Nadai has said the company came out of a simple frustration: after college, friends settle down, move away, have kids, or just can’t line up calendars anymore. He and co-founder Fabio Bin had tried other group-travel products for solo travelers, but felt the trips were missing something important—people were traveling together without really connecting. That’s what pushed the team to build group travel around age proximity and shared references. Social chemistry mattered more than just destinations.

Founder-market fit

Paolo De Nadai didn’t come into this cold. He founded ScuolaZoo at 19, turning it into a youth media brand in Italy, and later launched OneDay Group in 2012, a company built around products and communities for younger audiences. That background matters here. WeRoad’s real edge isn’t classic travel operations; it’s understanding how Millennials and Gen Z discover experiences and join communities. It also understands how they build identity around shared moments. Erika De Santi stayed close to the operating side of that model and is listed among the company’s senior leaders as co-founder and managing director.

Traction before the U.S. push

The execution so far is strong enough to make the U.S. gamble credible. WeRoad generated €130 million in revenue in 2025, up 30% year over year, and took more than 100,000 travelers on trips in that year alone. Since launch, it has served more than 300,000 customers across more than 1,000 itineraries. It now works with over 4,000 group leaders globally. Roughly 60% of travelers go on to book another trip. That’s the kind of repeat behavior investors care about.

WeMeet added another useful signal. In 2025, the app and events business brought in more than 50,000 attendees across 35 cities and hit 150,000 downloads. That doesn’t make it a breakout consumer app yet. But it does show that WeRoad can get people to show up offline before asking them to commit to a 10-day international trip. It’s a clever way to reduce trust friction.

Inside the $58 million round

The new round is a Series C worth $58 million, led by Airbnb, with existing investors including H14 also taking part. That takes WeRoad’s total funding to about $100 million. The company had previously announced a €18 million Series B in late 2023, so this isn’t a sudden spike out of nowhere. It’s a follow-on bet after a couple of years of growth.

The money is earmarked for WeRoad’s first major expansion outside Europe, starting in the U.S. and specifically in Austin. The plan isn’t to blast into every American city at once. It’s to seed a few local communities and recruit coordinators. It also plans to host in-person events and build partnerships before scaling harder. That’s slower. But it’s also a lot less reckless than pretending a social-travel brand can launch nationally from day 1.

How WeRoad compares with rivals

WeRoad sits in an odd but interesting middle ground. On one side, there are old-school escorted tours and youth package-travel brands that are really selling itinerary convenience. On the other, there are friendship and event startups like Timeleft, 222, and Pie that monetize dinners, clubs, and local hangouts. WeRoad blends both ideas. It sells paid travel, but also designs for community before, during, and now outside the trip itself.

That’s the real positioning. The company doesn’t use destination experts as the hero product. It uses group leaders closer in age to travelers and organizes early-trip activities to break the ice. Now it plugs WeMeet into the funnel too. So the product isn’t just “a trip to Japan” or “a ski week.” It’s a structured way to turn a bunch of strangers into a temporary social unit—and maybe a lasting one.

Why does the WeRoad Series C matter?

Airbnb leading this round is the loudest signal in the story. It suggests a big travel platform sees value in companies that don’t just help people book places, but help them belong somewhere. That’s a different thesis from the last decade of travel tech, which mostly optimized search, price comparison, and inventory.

For WeRoad, the bigger point is operational. The company now has the capital to test whether its European playbook—community-led trips, age-matched groups, and local event seeding—can survive contact with the U.S. market. And because it’s entering through Austin with both trips and WeMeet, management is trying to build a city-level social engine, not just buy traffic and hope strangers trust each other.

There’s also a more skeptical read, and it’s fair. “Loneliness” has become a startup pitch category of its own, and plenty of companies can create buzz around events without building durable economics. WeRoad’s advantage is that travel tickets are high-value transactions. If local meetups help fill those trips more efficiently, this turns from a feel-good narrative into a serious consumer business.

How big is the group travel market?

It’s not a niche. IMARC pegs the global adventure tourism market at $552.6 billion in 2025, and Grand View Research puts the group segment of adventure tourism alone at about $87.9 billion in 2025, with a path to roughly $308.2 billion by 2033. Those are broad numbers, but they show why investors will keep backing companies that can capture even a tiny slice of organized experience-led travel.

The timing also makes sense. Younger travelers increasingly want experiences that are social, flexible, and easy to share, and market researchers are tying sector growth to social-media-driven travel discovery and direct booking behavior. That doesn’t automatically mean every “IRL economy” startup wins. But it does help explain why a company built around group identity, not just trip logistics, looks more relevant in 2026 than it would have a decade ago.

What happens next for WeRoad in the U.S.?

The next thing to watch isn’t just whether WeRoad can sell trips in America. It’s whether Austin meetups turn into a repeatable acquisition engine for those trips. If that loop works—local event, community trust, group booking, repeat purchase—the WeRoad Series C could look like one of the smarter travel bets of the year. If it doesn’t, this starts to look like a very expensive experiment in monetized friendship.

Read how Cognition raised more than $1B for Devin AI, an autonomous coding agent built to help enterprise teams plan, write, test, and manage software development workflows.

FAQ

– What was WeRoad’s latest funding round?  

 WeRoad’s latest round was a $58 million Series C led by Airbnb, announced on May 27, 2026. The raise brings total funding to about $100 million and is meant to finance the company’s first expansion outside Europe, starting in Austin.

– How does WeRoad work for solo travelers?  

 WeRoad works by placing travelers into small age-aligned groups built around a shared trip style, then assigning a group leader who handles logistics and starts a WhatsApp chat before departure. Travelers can browse trips by vibe and activity level, preview departure groups before booking, and join itineraries that usually run for 10 to 12 days.

– Who founded WeRoad?  

 WeRoad was founded in 2017 by Paolo De Nadai, Fabio Bin, and Erika De Santi. De Nadai brought unusually strong founder-market fit because he had already built ScuolaZoo and then OneDay Group, both aimed at younger audiences and community-led products.

– Is WeRoad a travel company or a social platform?  

 It’s both, and that’s the whole point of the business. WeRoad makes money from group travel, but it designs the product around social connection and now extends that idea into local meetups through WeMeet, which hosted 50,000 attendees across 35 cities in 2025.

Cognition builds Devin, an autonomous AI software engineer for enterprise development teams. Its new Devin AI funding round — more than $1 billion announced on Wednesday, May 27, 2026 — comes as companies try to cut engineering backlogs without hiring endlessly. Founded in 2023 by Scott Wu, Steven Hao, and Walden Yan, the San Francisco startup is now being valued like a company that could define a product category, not just ride a hype cycle. Investors are betting there’s still room for an independent AI coding company even while model giants crowd the market.

What does Devin AI actually do?

Devin takes a software task in plain English, inspects a codebase, sketches a plan, writes code, runs commands, tests its own work, and hands the result back inside a developer workflow. In practice, it starts by searching the repo for relevant files and snippets. Then it produces an initial assessment with findings and implementation questions before moving into a more detailed plan a team can approve or edit.

The product isn’t just a chat box. Devin gives users an embedded IDE, a terminal, and a browser view. Teams can watch what it’s doing in real time, jump in when needed, or take over directly. That matters. The difference between a flashy demo and something engineers will trust usually comes down to visibility and control.

It also goes past single-ticket work. Devin can spin up managed child sessions in parallel, each in its own isolated VM, to handle chunks of a larger migration or code change. It can analyze earlier sessions and turn successful work into reusable playbooks. It also maintains an internal knowledge base and schedules recurring jobs like nightly checks or routine maintenance.

It plugs into the systems enterprise teams already use, including GitHub, GitLab, Bitbucket, Jira, Slack, and Microsoft Teams. So the pitch isn’t “replace your engineers.” It’s closer to “give your engineers another worker that can actually execute.”

Who built Cognition and Devin AI?

The founding story

Cognition started in 2023 with Scott Wu, Steven Hao, and Walden Yan. All 3 founders came out of elite competitive programming circles, which helps explain why the company went straight after autonomous software engineering instead of building yet another thin wrapper around a frontier model. They weren’t chasing a generic AI app. They were chasing one of the hardest agent problems they could find.

Why these founders had real market fit

Wu, Cognition’s CEO, had already built at startup scale before Devin. He previously co-founded Lunchclub and made Forbes’ 2020 30 Under 30 list. He’s also known in programming circles for a string of top-tier competitive results, including 3 International Olympiad in Informatics gold medals and a third-place finish in Google Code Jam. That mix — consumer startup execution plus deep technical credibility — is rare, and VCs tend to pay up for it.

Hao, the CTO, brought applied AI infrastructure experience from Scale AI, where he worked as a senior engineer before starting Cognition. Yan, Cognition’s CPO, previously co-founded DeepReason. Together, the resume starts to make sense: one founder with startup-building reps and one with hard production AI experience. The third had prior founder scar tissue.

Traction and the new round

The company’s customer list now includes Mercedes-Benz, NASA, Goldman Sachs, and Santander. Cognition also disclosed a $492 million annualized revenue run-rate, with enterprise usage of Devin growing 50% month over month for the last 6 months. For a company this young, that’s not normal. It changes a funding round from belief to aggressive extrapolation.

That helps explain the price. Cognition raised more than $1 billion at a $25 billion pre-money valuation and a $26 billion post-money valuation. The round follows a $400 million raise in September 2025 that valued the company at $10.2 billion post-money, so the markup in just 8 months is huge.

Lux Capital, General Catalyst, and 8VC led the new financing. Existing backers including Elad Gil, Soma Capital, Omri Casspi, and Founders Fund also participated, alongside new investors such as Ribbit Capital, Atreides, and Layer Global.

How Cognition stacks up against Codex, Claude Code, Cursor, and Jules

This part matters. A year ago, it looked like model companies would own AI coding end to end. OpenAI’s Codex is a cloud-based software engineering agent that can run many tasks in parallel. Anthropic’s Claude Code lives closer to the terminal and developer tooling stack, with strong workflow automation and MCP-based connectivity. Cursor has become the most visible AI-native editor. It pushes agents across desktop, web, mobile, Slack, and GitHub. Google’s Jules is another asynchronous coding agent built to read code, fix bugs, and work in a cloud environment.

Cognition’s bet is narrower and bolder. Rather than being the general model maker or the everyday editor first, it’s trying to own the “AI coworker” layer for software teams — the part where work gets scoped, delegated, executed, checked, and repeated. It also picked up the remaining pieces of Windsurf in 2025 after Google’s acqui-hire move, which gave it more product and talent leverage right as this market got crowded.

Why does this Devin AI funding round matter?

This isn’t just another giant AI valuation stapled onto a nice demo.

Cognition is being funded like a company that could become a core enterprise software vendor. That changes the questions customers, rivals, and investors will ask next. The issue is no longer whether teams will try autonomous coding agents. They already are. The issue is which products become durable enough for regulated, high-stakes organizations to trust with bigger chunks of production work.

The fresh capital should give Cognition room to harden Devin where enterprise buyers care most — reliability and permissions. Observability, workflow depth, security, and support matter too. That’s also where this round feels less like consumer AI theater and more like infrastructure building. Big banks and industrial companies don’t buy tools because they’re fun. They buy them because the tooling fits how software gets shipped.

Investors are clearly backing that thesis. They’re not just financing model access. They’re financing the application layer that sits on top of those models and turns raw capability into repeatable engineering output.

How big is the AI coding agents market?

Pretty big already. The global AI code assistants market was estimated at $8.5 billion in 2025 and is projected to reach about $42.9 billion by 2033, with a 22.5% compound annual growth rate from 2026 through 2033. That kind of curve is why capital keeps flooding into code-generation and agent startups.

But adoption isn’t the same as trust. Stack Overflow’s 2025 developer survey found a widening trust gap around AI tools, with 46% of developers saying they don’t trust the accuracy of AI output. That’s good context for Cognition. It suggests the winners won’t be the loudest products. They’ll be the ones that can show their work and fit into existing engineering systems. They also need to give teams a sane way to supervise the agent.

That’s why the timing works. The market is large enough to matter, crowded enough to be brutal, and skeptical enough that enterprise execution still counts for a lot.

What should investors and customers watch next?

Cognition has already won the hardest thing to fake — attention from serious enterprise buyers and serious capital at the same time.

Now it has to prove that autonomous coding agents can hold up after the demo, after procurement, and after security review. That’s the next test for Devin AI funding as a story. If Cognition keeps converting experimental use into standard enterprise workflow, this round will look expensive only in hindsight.

Read how Stord raised $250M at a $3B valuation to build independent commerce infrastructure that helps e-commerce brands manage warehouses, fulfillment, and shipping without relying on Amazon.

FAQ

– What is the latest Cognition funding round for Devin AI? 

 Cognition raised more than $1 billion on May 27, 2026. The financing valued the company at $25 billion pre-money and $26 billion post-money, a massive jump from the $10.2 billion post-money valuation attached to its September 2025 round.

– How does Devin AI work for software teams? 

 Devin takes a task in natural language, searches the relevant codebase, proposes a plan, and then writes, runs, and tests code inside its own working environment. Teams can watch the process through an IDE, terminal, and browser view. They can approve plans before execution and let Devin split large jobs into parallel managed sessions.

– Who founded Cognition? 

 Cognition was founded in 2023 by Scott Wu, Steven Hao, and Walden Yan. Wu previously co-founded Lunchclub, Hao worked at Scale AI, and Yan earlier built DeepReason, which is a big reason the company has looked unusually credible in AI software engineering from day 1.

– Is Cognition in the AI coding assistant market or something broader? 

 It sits in the AI coding assistant market, but its product ambition is broader than autocomplete or code suggestions. Cognition is aiming at the emerging AI coding agent category, where software tasks are not just assisted but planned, executed, and managed asynchronously across enterprise workflows.

Stord runs warehouses, fulfillment operations, and commerce software for e-commerce brands that don’t want Amazon owning the customer relationship. The Stord funding round announced on May 26, 2026 brought in $250 million at a $3 billion valuation, with Strike Capital leading and Kleiner Perkins, Founders Fund, Franklin Templeton, Baillie Gifford, G Squared, and Bond joining in. For a lot of online brands, the problem is simple: shipping fast is hard when inventory, carriers, and warehouse systems all live in different places. Sean Henry and Jacob Boudreau founded Stord in Atlanta in 2015. This round shows investors still think there’s room for an independent commerce infrastructure winner.

What does Stord actually do for brands?

At the basic level, Stord sells a mix of physical logistics and software. A brand can use Stord’s fulfillment network to store inventory and route orders. It also manages warehouse activity and pushes shipments out across DTC and B2B channels without stitching together a pile of separate 3PLs, dashboards, and carrier tools. That’s why the company has long pitched itself as an alternative to Amazon-style fulfillment for brands that still want to control the customer experience themselves.

The software side is more specific than the source article makes clear. Stord One Warehouse is a cloud warehouse management system that helps brands manage inventory and fulfill orders. It uses mobile scanning and automated workflows to replace manual work around carrier selection, inventory tracking, work orders, and cycle counts. In plain English: fewer spreadsheets, fewer paper processes, and less guessing about what’s sitting where.

The newer layer is AI. In March 2026, Stord rolled out StordAI assistants built around Chat, Search, and Feed. Chat lets customers ask plain-language questions about orders, inventory, shipment delays, routing, compliance rules, and forecast risk. Search works as one bar to pull up an order lifecycle or SKU-level inventory picture across systems. Feed pushes alerts about disruptions, demand shifts, and inventory risk before a human has to go hunting for them.

That matters because the customer experience changes a lot when operations aren’t buried in five tools. Before, an ops team might have to export reports and cross-check carrier data. It might also have to schedule internal calls just to answer why a shipment slipped. After, Stord’s pitch is that the answer should be available in seconds, inside the same platform that’s already running the fulfillment network. It’s an attempt to turn warehouse and delivery software into something operators can actually use during the workday, not just after the fact.

Who founded Stord before this funding round?

The founding story

Stord started in 2015 with Sean Henry as CEO and Jacob Boudreau as CTO. Henry’s own description of the origin is straightforward: he’d seen how fragmented 3PL operations were while running e-commerce businesses and while working in supply chain optimization at an automotive manufacturer in Germany, so he and Boudreau set out to make supply chains a competitive advantage instead of a tax on growth. That thesis still runs through the business now.

Why Sean Henry and Jacob Boudreau fit this market

Henry attended Georgia Tech before founding Stord and later became a Thiel Fellow. Boudreau attended Arizona State University before co-founding the company, was part of the 2016 Dynamo Accelerator cohort, and was later recognized as a Kairos K50 founder and a Forbes 30 Under 30 honoree alongside Henry. Neither founder came out of a giant legacy logistics company. That’s partly the point. They attacked the market like software builders who were annoyed by how clunky logistics still was.

The numbers behind Stord’s run

Stord hit unicorn status in 2021, then made it through the uglier funding stretch that followed. By May 16, 2025, the company had grown contracted revenue 10x since 2021, reached sustained profitability in 2024, powered more than $6 billion in commerce, and delivered over 30 million packages to roughly 11.5% of U.S. homes in 2024. It had also expanded to 11 fulfillment nodes across 13 buildings and shipped billions of units. Those are real operating numbers, not just “we’re growing fast” filler.

The funding history

The new round doubles Stord’s valuation from the prior year. In 2025, the company raised more than $200 million at a $1.5 billion valuation, also led by Strike Capital. With the May 2026 financing, total capital raised is now about $775 million. That’s a lot of money for a business that has to blend software economics with the messier reality of warehouses, transport, and labor.

Where Stord sits against Amazon and other 3PLs

The direct comparison is Amazon’s fulfillment machine. Amazon gives sellers speed, density, and reach, but it also pulls brands deeper into Amazon’s orbit. Stord’s pitch is the opposite: brands get national fulfillment and commerce software while keeping the customer relationship, which is why the company likes the “anti-Amazon” framing and the promise of giving merchants “the speed to compete.”

There are other rivals too, just in different shapes. ShipBob is a clear tech-enabled fulfillment competitor, and legacy 3PLs still handle huge volumes for brands that don’t want to build in-house. Stord stands out by combining first-party operations and a partner network. It also brings together B2B and DTC fulfillment, warehouse software, and now an AI layer in one product stack. Investors are betting that this combined model is harder to copy than another software dashboard or warehouse roll-up.

Why the Stord funding round matters now

This round matters because it isn’t just another valuation headline. It comes after the company survived the post-2021 pullback, kept scaling, and then layered AI on top of physical operations already in market. That’s a much stronger story than a startup promising future logistics magic without the warehouses, shipping volume, or customer data to back it up.

It also sharpens Stord’s roadmap. The company has already moved beyond an asset-light fulfillment marketplace into a broader commerce-enablement platform, and this capital gives it more room to expand the software side while keeping the network fast and cheap enough to matter. Frankly, that balance is the whole bet. If Stord were only software, Amazon could shrug it off. If it were only warehouses, margins would be tougher and the moat would look thinner.

Timing helps too. Google highlighted Stord at Cloud Next in April 2026, right as the company was pushing its AI interface deeper into daily operations. That doesn’t guarantee anything. But it shows Stord is getting attention for more than just moving boxes around.

How big is the e-commerce fulfillment market?

It’s big enough to justify this kind of capital. The global e-commerce fulfillment services market was estimated at $123.7 billion in 2024 and is projected to reach $272.1 billion by 2030, which implies a 14.2% CAGR from 2025 to 2030. North America accounted for 24% of that market in 2024, and the U.S. is expected to lead globally by 2030.

The trend line behind that growth is pretty simple. Consumers expect fast delivery, accurate inventory, and better visibility after checkout, but most brands still operate on fragmented systems. Stord’s own AI launch notes point to why that gap matters: 58% of consumers want estimated delivery dates, only 1% of brands provide them, and 25% to 35% of support tickets are still basic “Where is my order?” questions. That’s not a small inefficiency. It’s a structural opening for software-heavy logistics providers.

So this isn’t just a warehouse story. It’s about commerce operations being rebuilt around speed, data visibility, and automation. The old model still exists: one 3PL here, one carrier there, one spreadsheet everywhere. But it looks a lot weaker when brands need two-day reach, tighter margins, and instant answers on inventory and delivery performance.

The takeaway on Stord funding

The Stord funding round looks like a vote for something more specific than “AI plus logistics.” It’s a bet that independent brands still want a non-Amazon path to fast fulfillment, and that the winning product in this market won’t be software alone or warehouses alone, but a stubborn combination of both.

Execution is what matters next. If Stord can keep translating shipment data and inventory data into tools operators actually trust, the $3 billion valuation will look a lot more grounded. If the AI layer turns into window dressing, people will notice fast.

Read how Flexprice raised a $1.5M seed round co-led by Shastra VC and Anupam Mittal to build AI-native usage billing infrastructure for SaaS and AI companies managing complex pricing, metering, and invoicing workflows.

FAQ

– What happened in the latest Stord funding round? 

 Stord raised $250 million on May 26, 2026 at a $3 billion valuation. Strike Capital led the round, and the investor list included Kleiner Perkins, Founders Fund, Franklin Templeton, Baillie Gifford, G Squared, and Bond, bringing total funding to about $775 million.

– How does Stord actually work for e-commerce brands? 

 Stord combines a fulfillment network with commerce software so brands can manage inventory, warehouse activity, order routing, and delivery operations in one system. Its newer AI tools add natural-language order lookup, SKU and inventory search, and proactive alerts about delays, demand changes, and stock risk.

– Who founded Stord? 

 Stord was founded in 2015 by Sean Henry and Jacob Boudreau, who still serve as CEO and CTO. Henry came into the company after hands-on e-commerce work and supply chain optimization experience in Germany, while Boudreau built the technical side and had early founder credentials through Dynamo and Kairos before Stord scaled into a unicorn.

– Is Stord a logistics company or a software company? 

 It’s both, and that hybrid model is the point. Stord operates in e-commerce fulfillment and 3PL software, selling warehouse and transportation execution alongside tools like warehouse management, order visibility, and AI-driven operations support for omnichannel brands.

Flexprice builds usage billing infrastructure for AI and SaaS companies that need to meter product activity, turn that usage into pricing, and send invoices without stitching together a pile of internal tools. The New Delhi startup has raised a $1.5 million seed round co-led by Shastra VC and Shaadi.com founder Anupam Mittal, with existing backer TDV Partners also joining in. Billing is becoming a real product problem for AI companies, not a back-office chore. Flexprice was founded in 2024 by Manish Choudhary, Koshima Satija, and Nikhil Mishra, and it already operates across New Delhi, San Francisco, and Bengaluru.

What is Flexprice and how does its billing infrastructure work?

Flexprice is an open-source platform for metering, billing, and feature management. In plain English: a customer defines what should be tracked — API calls, tokens, compute time, seats, credits, or some other usage signal — sends those events into Flexprice, connects them to a pricing plan, and the platform calculates charges and generates invoices automatically. The product docs lay out that flow directly: create a metered feature, send events, validate them, connect them to billing, then let invoicing run.

That’s more than a billing wrapper. Flexprice’s architecture is split into composable layers for usage metering, pricing, subscriptions, entitlements, and invoicing. Teams can use the whole stack or wire in only the pieces they need. The docs also show support for real-time event ingestion, usage analytics, and feature limits. Useful for AI products that want to bill on consumption while also turning features on or off based on plan or credits.

The product also goes beyond classic subscriptions. It supports seat-based, usage-based, and hybrid pricing. It also supports credit grants, auto top-ups, custom invoice logic, add-ons, bundles, price localization and ships SDKs for JavaScript, Python, and Go, and offers self-hosting for teams that don’t want vendor lock-in or black-box billing logic.

For customers, the before-and-after is pretty obvious. Before Flexprice, teams often end up writing custom code for proration, limits, credits, taxes, and invoice reconciliation. After integration, the product handles usage aggregation in real time. It previews invoices, manages invoice states, and gives finance teams a cleaner audit trail. That’s why the company pitches itself as revenue infrastructure, not just another payment plugin.

Who founded Flexprice?

The founding story behind Flexprice funding

Flexprice was founded in 2024 by Manish Choudhary, now CEO, Koshima Satija, now COO, and Nikhil Mishra, now CTO. The company’s origin story is unusually concrete: Satija has written that the spark came while Choudhary was at AI photo-editing company Aftershoot and struggling with localized pricing during international expansion. The founders kept running into the same issue — pricing ideas were easy to sketch, but ugly to implement once geography, usage, credits, and invoicing rules showed up.

Why these founders make sense for this category

Choudhary has said he previously worked a full-time job while running a consultancy that became an agency helping companies figure out pricing. That matters because Flexprice sits at the intersection of monetization strategy and product plumbing. He wasn’t coming at this as a generic founder chasing an AI trend. He was already dealing with the messy part of how companies actually charge.

Mishra brings the engineering side. Choudhary said in a Reddit AMA that Mishra had been engineering head at WizCommerce and Zomato, which gives him credibility on the systems side of high-scale data and operational infrastructure. In the same AMA, Satija and Choudhary described themselves as product and business operators from AI and consumer companies. Satija’s own writing frames Flexprice as a product-manager-plus-tech-lead insight — business requirements meeting engineering pain.

Early execution signals

The product is live, not a deck. Flexprice launched on Product Hunt on April 6, 2025 and finished as Product of the Day with 500+ upvotes and 50+ sign-ups. The company has a community of 300+ builders on Slack, and Choudhary wrote earlier this year that Flexprice is used by companies including Krutrim and Simplismart, has 3,500+ GitHub developers around the project, and processes more than 20 billion API requests a month. Those are founder-reported numbers. Read them as early traction signals rather than audited metrics.

Flexprice funding round details

This new $1.5 million round is a seed financing co-led by Shastra VC and Anupam Mittal, with TDV Partners returning. Before that, Choudhary disclosed an earlier $500,000 round led by TDV Partners with support from angel investors. The new money is meant for expansion across the US and Europe, plus new product development. Flexprice is headquartered in New Delhi and already has teams in San Francisco and Bengaluru.

Avijeet Alagathi of Shastra VC said the appeal was the team’s “engineering vision” and the fact that Flexprice is an open-source, real-time billing platform built for modern AI products while still fitting into existing systems. TDV’s Ujwal Sutaria leaned on the same point — the open-source approach and strong technical execution. That’s a pretty clear investor read. Not just “billing is growing,” but “billing built for AI-native products is underbuilt.”

Flexprice funding round details

Flexprice isn’t entering an empty category. Metronome has become one of the best-known names in usage-based billing and says OpenAI uses it for scalable billing infrastructure. Orb pitches a similarly modern stack for usage-based, seat-based, and hybrid billing. It also supports prepaid credits and complex enterprise contracts. Lago comes from the open-source side and positions itself as software for metering and usage-based billing with payment-provider connections.

So where does Flexprice try to wedge in? Mostly in three places. First, open source and self-hosting, which matters for teams that don’t want core revenue logic hidden inside a vendor box. Second, a tighter AI-native framing — tokens, credits, limits, and rapidly changing pricing models are native objects in the product. Third, a broader product surface that mixes billing with feature management and entitlements, which legacy subscription tools often treat as somebody else’s problem. Its real competition, honestly, isn’t only Metronome or Orb. It’s also homegrown billing stacks, spreadsheets, and subscription-era tools that were never built for usage-heavy products.

Why Flexprice funding matters for AI startups?

A $1.5 million seed round isn’t huge by AI-infra standards. But for Flexprice, it looks less like a vanity round and more like fuel for distribution and product breadth.

The company already has the bones of a real platform — metering, credits, feature controls, invoicing, self-hosting. What it didn’t have, at least publicly, was the scale to push harder in the US and Europe while still building toward Choudhary’s stated goal of “full revenue automation.” That phrase deserves attention. It suggests Flexprice wants to move upstream from billing events and downstream toward recognized revenue. That’s a much bigger ambition than “we help you send invoices.”

It also gives customers a signal. If you’re an AI startup deciding whether to keep hacking together billing internally, outside money from Shastra VC, Mittal, and TDV tells you this isn’t a side project anymore. And if the founders keep shipping, the open-source angle could become a practical buying reason, not just a brand story.

How big is the usage billing infrastructure market?

The macro tailwind is real. Grand View Research estimates the global software segment of the subscription billing management market was worth about $4.8 billion in 2024 and could reach roughly $11.7 billion by 2030. It also expects India to post the fastest growth rate in that period, which is a useful backdrop for a company built in India but selling into global software markets.

The adoption curve is moving too. OpenView’s SaaS benchmarks work, as summarized by TechCrunch, found that 61% of SaaS companies used usage-based pricing in some form in 2022. OpenView also said startups offering a usage-based model across its customer base rose by 30%. That doesn’t mean every SaaS company wants pure consumption pricing. A lot of them don’t. But it does mean billing logic is getting more complex, and complexity creates room for new infrastructure vendors.

AI makes that sharper. Seat-based pricing was fine when software value mapped cleanly to headcount. It breaks down when one user can trigger millions of tokens, GPU-heavy workflows, or autonomous tasks. The winners in this category will be the vendors that let product teams change pricing fast without breaking finance. That’s the market Flexprice is chasing.

Should buyers watch Flexprice now?

Probably yes — with some healthy skepticism.

Flexprice has the right shape for the current moment: open-source roots, AI-native billing logic, and founders who seem to understand that pricing changes are easy to announce and painful to operationalize. But this is still an early company in a category with serious incumbents and strong developer-first rivals. The next thing to watch isn’t another funding headline. It’s whether Flexprice can turn this seed round into deeper enterprise adoption in the US and Europe while keeping its usage billing infrastructure flexible enough for the weird pricing models AI companies keep inventing.

Read how Fairdeal.Market raised $15M in Series A funding led by Bertelsmann India Investments to help kirana stores restock FMCG inventory in under 60 minutes through its B2B quick commerce and dark-store network.

FAQ about Flexprice

– What funding did Flexprice raise?  

 Flexprice raised a $1.5 million seed round announced on May 26, 2026. Shastra VC and Anupam Mittal co-led the round, with TDV Partners also participating after backing the company earlier.

– How does Flexprice actually work?  

 Flexprice works by letting a company define what usage it wants to track, send those events into the platform, connect them to pricing plans, and let the system calculate charges and generate invoices. It also handles credits, feature limits, subscriptions, and hybrid pricing models. That makes it closer to monetization infrastructure than a simple invoicing tool.

– Who founded Flexprice?  

 Flexprice was founded in 2024 by Manish Choudhary, Koshima Satija, and Nikhil Mishra. Choudhary is CEO, Satija is COO, and Mishra is CTO; the founders tie together pricing strategy, product experience, and engineering depth from earlier work across AI, consumer, WizCommerce, Zomato, and startup operating roles.

– What market is Flexprice in?  

 Flexprice sits in the subscription and usage-based billing software category, with a strong focus on AI-native revenue infrastructure. It’s a growing market: Grand View Research pegs the software portion of subscription billing management at about $4.8 billion in 2024, while SaaS adoption of usage-based pricing has already become common enough that more than half of companies use it in some form.

Fairdeal.Market is a B2B quick commerce platform that supplies kirana stores and other neighborhood retailers with fast-turn FMCG inventory across Delhi NCR. The problem it’s chasing is simple and old: small retailers still lose time, margin, and sales when sourcing stock is fragmented and slow. Fairdeal Market funding just got a big boost, with the Gurugram-based startup raising $15 million in Series A capital led by Bertelsmann India Investments. WaterBridge Ventures and Incubate Asia Fund also joined the round. Co-founders Prateek Bansal and Yash Bansal started the company in 2022, and this round gives them more room to scale a business that’s already serving a large retailer base in a brutally execution-heavy category.

What does Fairdeal Market do?

Fairdeal.Market works like a hyperlocal wholesale layer for retailers. A shopkeeper places an order for FMCG inventory through the platform, chooses from a broad catalogue across everyday categories, gets wholesale pricing visibility, and receives delivery in under 60 minutes through Fairdeal’s dark-store and last-mile network. The pitch isn’t fancy. It’s speed, better sourcing, and fewer stock-outs for stores that can’t afford to wait half a day for replenishment.

That workflow matters because the platform isn’t just listing products. It compresses the retailer’s entire reorder cycle. Retailers can track spending and reorder in seconds, which is a real operational shift for stores that used to call multiple distributors or make mandi runs just to fill routine gaps on shelves. The product is a one-stop wholesale app for categories ranging from cold drinks and snacks to dairy, personal care, cleaning essentials, and staples.

The “quick commerce” label here is slightly different from consumer grocery apps. Fairdeal is built for stores, not households. That means the customer experience is less about impulse convenience and more about working capital rotation. One retailer testimonial says the shift from bulk buys of ₹30,000-35,000 to smaller daily orders of ₹5,000-6,000 improved cash flow. Another says ordering 3-4 times a day replaced daily mandi trips. That’s the product story. Smaller purchases, faster refill cycles, and less dead inventory sitting in the shop.

There’s also a supply-side angle. Fairdeal pitches itself to brands as a last-mile distribution channel into local retail. More than 25 brands expanded retailer reach by 38% in 4 months through its network. So the platform isn’t only serving kiranas. It’s trying to become a dense urban distribution pipe for FMCG brands that want better store-level coverage without relying entirely on older distributor structures.

Who founded Fairdeal Market?

The company started with a narrow, practical thesis

Fairdeal was founded in 2022 by Prateek Bansal and Yash Bansal. The company is based in Gurugram and operates as FDM Digital Solutions Pvt Ltd. Prateek Bansal is listed as co-founder and CEO, while Yash Bansal is listed as co-founder and CIO. That job split tells you a lot already. One side is on business buildout, the other on systems and operating intelligence.

The founding idea is pretty grounded: don’t try to digitize all of Indian retail at once. Start with dense urban clusters, build dark stores, focus on frequent-fill retail categories, and win on speed. That’s less glamorous than a giant national marketplace story. But it’s a lot easier to believe.

Early traction is real, even if the target is aggressive

The company delivers more than 1,000 SKUs to retailers across Delhi NCR within 60 minutes. Over the last 6 months, it scaled to more than 20,000 active retailers in the region, while maintaining customer retention above 80%. It now wants to expand that retailer base to more than 100,000 within the current financial year. That ambition is huge. And the execution risk is obvious.

There are a few other signals that help. Fairdeal’s company profile lists employee strength in the 21-40 range. The platform also showcases a growing list of brand partners across packaged foods, beverages, and household products. For a company founded in 2022, that suggests it’s past the “PowerPoint startup” stage and firmly in the ops-heavy build phase.

The fundraising path got bigger, fast

This Series A round brought in $15 million, or about ₹142.8 crore, led by Bertelsmann India Investments, with participation from WaterBridge Ventures and Incubate Asia Fund. Before that, Fairdeal raised $3 million in a pre-Series A round led by Incubate Fund Asia and WaterBridge Ventures, with angel investors participating in August last year.

The new money is supposed to do 4 things. Expand dark-store operations in dense urban clusters. Improve the company’s tech and data stack. Deepen retailer engagement. Strengthen last-mile delivery. That mix makes sense. A business like this dies if any one of those layers falls behind.

How Fairdeal compares with Udaan and Jumbotail

Fairdeal isn’t entering an empty market. Udaan is still the best-known eB2B name in India and closed a $114 million Series G round in June 2025 as it kept pushing toward a stronger balance sheet and IPO prep. Jumbotail raised $120 million in June 2025 and crossed the $1 billion valuation mark, while also broadening its reach through the Solv India combination. Those companies operate at much larger scale.

Fairdeal looks less like a broad national eB2B marketplace and more like a high-frequency, hyperlocal replenishment engine. Udaan and Jumbotail have spent years building multi-city distribution and broader category coverage. Fairdeal’s bet is tighter: dense urban geography, fast delivery windows, smaller order cycles, and a dark-store model tuned for retailers that need top-ups several times a day. That differentiation matters because the incumbent alternative for many kiranas still isn’t another app. It’s a patchwork of local distributors, phone calls, and physical wholesale runs.

Why Fairdeal Market funding matters

This round matters because it changes what Fairdeal can realistically attempt next. Until now, the company had shown it could get retailers to use the service and stick with it. But retention is one thing. Building a repeatable city-density machine is another. Dark stores, delivery routing, inventory placement, and retailer engagement all get expensive fast.

The investor list is interesting for the same reason. Bertelsmann India Investments didn’t back a slide deck here. It backed a logistics-heavy model that only works if demand density and operational discipline show up together. WaterBridge and Incubate Asia Fund returning also signals that earlier backers think the core behavior is holding up. Frequent wholesale ordering through a fast local network.

For retailers, the practical effect could be immediate. If Fairdeal uses the money well, shop owners should see deeper assortment and more reliable fill rates. Faster replenishment should follow across more pockets of Delhi NCR and nearby urban clusters. If it uses the money badly, the model gets exposed very quickly. This category is unforgiving like that.

How big is the kirana and B2B quick commerce market?

India’s retail backbone is still overwhelmingly offline. Redseer estimates that roughly 85%-90% of mass grocery retail in India continues to run through traditional trade, with around 13-14 million kirana stores across the country. That scale explains why startups keep chasing this segment even after plenty of eB2B companies learned the hard way that distribution economics can get ugly.

The structural trend helping companies like Fairdeal is density. Quick commerce and hyperlocal supply models work best where order frequency is high and basket sizes are small. Stores also need fast refill cycles. Dense urban micro-markets are exactly where those conditions exist. Fairdeal’s city-cluster approach lines up with that reality better than the old “expand everywhere” playbook that burned so much capital in Indian B2B commerce.

There’s another reason the timing feels right. Kiranas are still the default retail channel for a huge part of the country, but their sourcing behavior is getting more digital. That doesn’t mean distributor networks disappear. It means the best platforms can start taking share from the messy parts of the old system — stock gaps, slow fulfilment, opaque pricing, and wasted procurement time.

Conclusion

Fairdeal.Market hasn’t won anything yet. But this is the kind of startup worth watching because the problem is real and the operating model is concrete. Fairdeal Market funding gives the company enough firepower to test whether retailer quick commerce can scale beyond a promising Delhi NCR base. The next thing to watch is simple: can it turn dense-city traction into repeatable unit economics before bigger rivals crowd the lane?

Read how Hark raised over $700M in a massive Series A round to build an AI personal assistant platform and future hardware designed to act as a universal interface across the digital tools people already use.

Fairdeal Market funding FAQ

– What is the latest Fairdeal Market funding round? 

 Fairdeal.Market has raised $15 million in a Series A round. Bertelsmann India Investments led the round, and WaterBridge Ventures plus Incubate Asia Fund also participated, giving the startup fresh capital to expand its dark-store and delivery network.

– How does Fairdeal.Market work for kirana stores? 

 Fairdeal.Market lets retailers order FMCG inventory through a B2B quick commerce platform and receive deliveries in under 60 minutes. The service is built around high-frequency replenishment, so stores can buy smaller quantities more often instead of tying up cash in larger wholesale purchases.

– Who founded Fairdeal.Market? 

 Fairdeal.Market was founded in 2022 by Prateek Bansal and Yash Bansal. Prateek is listed as co-founder and CEO, and Yash is listed as co-founder and CIO, with the business headquartered in Gurugram and operating across Delhi NCR.

– Is Fairdeal.Market in B2B ecommerce or quick commerce? 

 It’s really both, but the sharper label is B2B quick commerce. The company sells wholesale FMCG inventory to retailers, yet its core differentiation is sub-60-minute replenishment through dark stores and last-mile delivery rather than a slower national marketplace model.

Hark is building an AI personal assistant platform and future hardware meant to become a universal interface for the digital tools people already use. Hark AI funding just landed at an eye-watering level: more than $700 million in a Series A round that values the company at $6 billion post-money. Founded in late 2025 by Brett Adcock, Hark is chasing a very simple problem that turns out to be brutally hard — most AI products still feel impressive in demos and awkward in normal life. That’s why this raise matters. It’s a bet that a consumer AI product people actually want might need huge capital, not just a clever model.

What is Hark and how will its AI assistant work?

Hark still hasn’t shown the full product, which is part of why the round is so striking. But the company has said enough to sketch the core idea: its first platform is due in summer 2026, and it will use agentic, multimodal models that remember who you are and what you say. Those models will work across the products and services you already use, so the assistant isn’t supposed to live inside one app. It’s supposed to sit above them.

In plain English, Hark is pitching a system that can take context from your ongoing digital life and keep track of preferences and history. Then it acts across services on your behalf. That means less jumping between apps and less repeating the same instructions. Less babysitting too. Hark hasn’t broken out exact features yet, but its public description points to a memory layer plus action-taking software, not just a chatbot with a nicer shell.

The hardware comes after that. Hark says a new class of AI-native devices will follow the software launch, built specifically to integrate with its own foundation models rather than bolting AI onto an old product category. That makes this feel closer to a full-stack consumer AI company than a typical app startup. It’s also the expensive part. And risky. Hardware graveyards are full of smart ideas that never solved comfort, trust, battery life, or simple social weirdness.

Who founded Hark and why are investors backing it?

The founding story

Brett Adcock launched Hark in late 2025 with $100 million of his own money. The pitch was ambitious from day 1: build an agentic AI system that works as a universal interface to the digital world, then pair it with dedicated hardware instead of stopping at software.

That sounds huge because it is. But it also fits Adcock’s pattern. He doesn’t really do modest categories.

Founder-market fit

Adcock is best known as the founder of Figure, the robotics company building general-purpose humanoids. Before that, he founded Archer, which went public at a $2.7 billion valuation, and Vettery, a machine learning-based hiring marketplace that exited for about $100 million. That doesn’t guarantee Hark works. It does explain why investors are willing to hand him a monster Series A before the product is fully out in the open. He’s built capital-heavy companies before. He knows how to recruit around hard engineering problems.

Hark’s design credibility also got a boost from Abidur Chowdhury, a former Apple designer who joined after working on Apple’s industrial design team from 2019. He later appeared in Apple’s iPhone Air presentation, which made his move stand out even inside Apple circles. At Hark, he’s now directing design for a company that clearly wants to treat interface and hardware as the main event, not an afterthought.

Early signals from the company

For a company this secretive, Hark has still revealed a few useful signals. The team has grown to around 70 people. The platform is entering beta. It plans to train its next generation of models on a new Nvidia B200 data center. That doesn’t tell you whether consumers will want the product. It does tell you this isn’t a two-slide concept with a nice video and no infrastructure behind it.

Chowdhury’s public framing is also telling. He argued that a lot of AI companies are building tools that help people make software, and that those products do work, but he hasn’t seen much that really helps “the normal person.” He contrasted Hark’s focus with companies leaning harder into coding assistance. That’s a sharp positioning move. Less developer utility, more mass-market interface.

Fundraising details

Parkway Venture Capital led the round and Nvidia, Align Ventures, AMD Ventures, ARK Invest, Brookfield, Greycroft, Intel Capital, Prime Movers Lab, Qualcomm Ventures, Salesforce Ventures, and Tamarack Global joined in. For a Series A, that’s absurdly large. It’s also a pretty specific cap table. You don’t bring in that many chip, platform, and strategic names unless you think compute, components, and distribution will matter early. Hark says the money will go toward hiring across hardware and product design. AI research too. It also plans to lock down compute and parts.

Competition and market positioning

Hark isn’t walking into an empty market. OpenAI bought Jony Ive’s io in May 2025 for $6.5 billion to build a family of AI devices, though reporting since then suggests that first hardware won’t ship until 2027. Meta already has consumer AI glasses in market. Google and its Android partners are pushing in too. And standalone AI gadget startups have already shown how ugly this can get — Humane’s AI Pin was effectively over after HP bought parts of the company in February 2025.

Hark’s angle is different enough to stand out. It wants its own models. It wants native hardware. And it wants a broad personal assistant for ordinary users, not a device built mainly to show off AI tricks or help developers write code faster. But there’s a giant catch. The company still hasn’t answered the nastiest question in this category: how do you gather enough context from a user’s life to be genuinely helpful without creeping out everyone around them? Chowdhury’s joking answer — “Sounds like that would make a great product” — was funny because it’s the whole challenge.

Why does Hark AI funding matter right now?

Because this round gives Hark permission to attempt two brutally expensive jobs at once.

One is frontier-ish AI product work. The other is consumer hardware. Most startups pick one. Hark is trying to do both, and that means hiring top researchers, industrial designers, product people, and supply-chain talent before revenue is anywhere close to certain. A $700 million Series A buys time for that.

It also changes the investor conversation. This isn’t just “AI assistant app gets funded.” It’s a signal that some big backers think the next must-have consumer AI product may need a full-stack approach. Models, memory, interface, chips, hardware, and design under one roof. That’s a much bigger swing than wrapping an API in a nice UI.

There’s a subtler point here too. Hark can stay weird for longer. It doesn’t have to rush into a compromised first device just to keep the lights on. That matters in consumer hardware, where the early version has a nasty habit of defining the whole company.

How big is the market for AI hardware and personal assistants?

The macro case is real, even if the winners are still messy. Grand View Research estimates the U.S. wearable AI market generated about $6.1 billion in 2023 and could reach roughly $39 billion by 2030, with a 30.4% CAGR from 2024 to 2030. Technavio forecasts the personal AI assistant market will expand by $12.36 billion from 2025 to 2030 at a 34.8% CAGR. Those are big numbers. More important, they point in the same direction: people are starting to expect AI to be ambient, persistent, and built into devices, not trapped in a single chat window.

That’s why so many companies are suddenly obsessed with interfaces. The software wave proved people will use AI. The next question is where it lives. On your phone? In your glasses? In an always-listening wearable? In something totally new? OpenAI’s move into hardware, Meta’s push with smart glasses, and Hark’s own full-stack bet all come from the same basic realization: the model matters, but the wrapper now matters a lot too.

What should you watch after Hark AI funding?

Hark AI funding is massive, but money doesn’t make a consumer habit by itself.

What matters next is simple. Does Hark’s summer launch show a real assistant with durable memory and useful actions, or just another polished demo? And when the hardware finally appears, does it solve the privacy and comfort problem better than the last wave of AI gadgets did?

Read how fragrance tech startup Patina raised $2M from Betaworks and True Ventures to build AI-powered scent molecules and turn fragrance creation into a programmable science platform.

FAQ: Hark AI funding

– What is the Hark AI funding round?  

 Hark raised more than $700 million in a Series A round at a $6 billion post-money valuation. Parkway Venture Capital led the deal, and the investor list included Nvidia, AMD Ventures, Qualcomm Ventures, Intel Capital, Salesforce Ventures, ARK Invest, and several others.

– How does Hark’s product work?  

 Hark is building an agentic, multimodal AI platform that remembers user context and works across existing products and services. The software is expected in summer 2026, and the company plans to follow it with AI-native hardware designed around Hark’s own models instead of adapting older device categories.

– Who founded Hark?  

 Brett Adcock founded Hark after earlier building Vettery, Archer, and Figure. That background matters because Archer reached a $2.7 billion IPO valuation and Vettery had roughly a $100 million exit, giving Adcock a track record with large, difficult company-building bets.

– Is Hark an AI hardware company or an AI assistant startup?  

 It’s both, at least by design. Hark is starting with a personal AI assistant platform in beta, but the long-term plan is a full-stack consumer AI business that combines proprietary models, memory, and dedicated hardware in one product stack.