Venture Capital & Startup Funding Roundup, June 24, 2026

It’s Wednesday, June 24, 2026, and today’s funding tape was not a broad-based venture rebound. It was a concentrated bet on production AI: systems that make real decisions, move money, route patients, govern agents, search the live web for machines instead of humans, and keep enterprise deployments from drifting into chaos.

Nine of the ten rounds in this report have a direct AI product, infrastructure, or governance angle, and the outlier, Caplight, still fits the same deeper pattern: investors are financing the rails needed for a market where companies stay private longer, and capital needs better pricing, liquidity, and workflow data.

What stands out even more than the sector mix is who is writing checks. Menlo Ventures, Goldman Sachs Alternatives, BlackRock, UBS Investment Bank, Aramco Ventures, Khosla Ventures, Norwest, and Speedinvest all appeared within the same 12-hour window. That is a signal founders should take seriously. This is not just venture capital chasing AI exposure. It is a mix of classic venture firms, strategics, and large institutions backing companies that sit near critical business workflows, regulated decision-making, or core market infrastructure.

The other theme is that investors are now rewarding proof, not just possibility. In this batch alone, companies pointed to 20x revenue growth, 95% automation in underwriting, 75% fewer AML false positives, 190 million patient interactions, 300 million medical records processed, more than 20 million agent actions a month, tens of millions of voice evaluations, and institutional clients managing trillions of dollars. The market is still paying for ambition, but only when ambition is paired with visible operating leverage.

The Macro Environment: Capital Is Chasing Production AI

The broader backdrop helps explain why today’s rounds feel so focused. Crunchbase reported that global startup funding hit a record $300 billion in the first quarter of 2026, up more than 150% year over year, but that surge has been highly concentrated rather than evenly distributed. PitchBook and NVCA described the U.S. venture market in the same quarter as “one for the record books,” while also warning that removing the five largest deals would slash deal value by 73.2%, and that 73.1% of new VC fund commitments went to just five firms. In other words, capital is plentiful at the top and scarce almost everywhere else.

That concentration is increasingly tied to AI, but not all AI is being funded the same way. Crunchbase found that U.S. companies have captured nearly 80% of global seed-to-growth startup financing so far in 2026, while Europe’s first quarter still improved to $17.6 billion, with AI accounting for more than half of regional venture funding for the first time. The implication for founders is blunt: the market still funds international startups, but the burden of proof is higher unless the company can claim a strong wedge such as regulation, sovereignty, or sector-specific depth.

Today’s rounds map almost perfectly onto that reality. In the U.S., investors favored application and governance layers attached to healthcare, finance, sales, private markets, and enterprise operations. In Europe, the standout signals were around sovereign AI infrastructure and regulated financial workflows rather than consumer apps. France’s recent €13 billion tech-sovereignty push under the Tibi initiative adds context here: European capital is increasingly being organized around strategic autonomy, not just startup growth for its own sake.

Public-market conditions also matter. PitchBook/NVCA noted that liquidity remains tight for much of the market, IPO registration activity has not meaningfully reopened, and fresh geopolitical and policy risks are weighing on investor appetite. That helps explain why one of today’s more revealing rounds is Caplight’s: when private companies stay private longer, the battle for better price discovery, secondary liquidity, and workflow data becomes a venture category in its own right.

The Funding Roundup

Assort Health raises $120 million in funding to scale AI agents for the patient journey

Assort Health’s round was the day’s clearest statement that healthcare AI is moving from narrow front-desk tools toward operating systems for patient access. The company says it has built the most widely used AI agents platform for the patient journey, spanning scheduling, intake, referrals, document processing, medication refills, lab requests, payments, and staff copilots. That breadth matters because healthcare buyers do not want another stack of disconnected point products; they want fewer vendors, tighter workflow coverage, and better memory across the patient journey.

Menlo Ventures led the $120 million Series C at a $1.2 billion valuation, giving Assort unicorn status and, more importantly, validation that investors now view healthcare administration as a large AI application market rather than a side bet. Assort says it has supported more than 190 million patient interactions, built on 62,000 care protocols and 1.6 million decision pathways, and grown revenue 20x in the last 15 months. That is the kind of operating data late-stage investors want to see before paying up in this environment.

Why does the round matter beyond the headline size? Because it sits at the intersection of two powerful forces: healthcare’s immense administrative burden and AI’s growing ability to automate domain-specific workflows. Assort argues that providers now spend nearly twice as much on administration as on direct patient care, and Menlo framed the investment as part of a thesis to back the strongest application-layer AI companies in each large sector. The underlying message is that AI in healthcare is becoming less about clever demos and more about replacing expensive, repetitive operating work.

Funding Details
Startup: Assort Health
Investors: Menlo Ventures, Lightspeed Venture Partners, Felicis, First Round Capital, Chemistry, Joe Montana, Tau Ventures, Quiet Capital
Amount Raised: $120 million
Total Raised: More than $222 million
Funding Stage: Series C
Funding Date: June 24, 2026
Headquarters: San Francisco, California, U.S.
Sector: Healthcare AI/patient access automation.

Taktile raises $110 million in funding to automate high-stakes financial decisions

New York-based FinTech startup Taktile has secured $110M in Series C funding, led by Goldman Sachs, to expand its AI decisioning for financial institutions. Taktile is one of the day’s most institutionally important financings because it targets one of the hardest problems in enterprise AI: using models in regulated financial decisions where errors create legal, compliance, and credit risk.

The company’s product lets banks and insurers combine AI agents, rules, relevant context, and human oversight to automate decisions across underwriting, claims, fraud, onboarding, and AML workflows. This is not generic productivity software. It is AI inside the decision loops that financial institutions actually care about.

Goldman Sachs Alternatives led the Series C, with support from Tiger Global, Index Ventures, Balderton Capital, Y Combinator, and Dig Ventures. That syndicate tells its own story. Growth capital is now willing to back AI companies that not only sell “co-pilots” but also sit directly in the operating core of banks and insurers. Taktile says it powers millions of decisions every day, that one major insurer is projecting more than $90 million in claims-processing efficiencies from use cases on the platform, and that customers have achieved 95% automation in B2B underwriting and 75% fewer AML false positives.

Strategically, Taktile shows where the AI-in-finance market is heading. Investors are no longer funding only horizontal tooling for model deployment; they are funding domain-native control layers that can withstand audits and operate within sector constraints. The company did not disclose its valuation, which is notable in itself. In a market still sensitive to inflated private marks, some of the strongest companies are choosing to emphasize deployment metrics and institutional logos over vanity numbers.

Funding Details
Startup: Taktile
Investors: Growth Equity at Goldman Sachs Alternatives, Balderton Capital, Index Ventures, Tiger Global, Y Combinator, Dig Ventures
Amount Raised: $110 million
Total Raised: $184 million
Funding Stage: Series C
Funding Date: June 24, 2026
Headquarters: New York City, with offices in Berlin and London
Sector: Financial services AI/decision automation.

xCures raises $46 million in funding to turn fragmented medical records into decision-ready clinical data

xCures sits in a less glamorous but highly defensible part of health AI: making messy medical records usable. Its pitch is that most healthcare data systems are still good at transporting records but poor at turning them into structured, decision-ready intelligence. That distinction is why this round matters. In health AI, the best application often depends on who owns the cleanest data layer, not who ships the flashiest assistant.

Innovius Capital led the Series B, and Crunchbase reported that the round values xCures at $127 million post-money, more than double its previous Series A valuation. The company says it has processed more than 300 million medical records from more than 550,000 U.S. healthcare locations and that its product has expanded from oncology roots into broader clinical use cases. That combination — sticky infrastructure plus visible scale — is what makes the story compelling.

For investors, xCures is a bet that healthcare’s next AI winners may look more like data plumbing than virtual doctors. The company’s customers include diagnostic labs, hospital networks, telehealth providers, and Medicare Advantage plans, which give it multiple paths to grow without relying on a single reimbursement model. In a market full of AI promises, xCures is selling a concrete output: cleaner data for real operational and clinical decisions.

Funding Details
Startup: xCures
Investors: Innovius Capital, iGrow, GKCC, Spring Mountain Capital, and existing investors
Amount Raised: $46 million
Total Raised: More than $76 million
Funding Stage: Series B
Funding Date: June 24, 2026
Headquarters: Oakland, California, U.S.
Sector: Health AI / clinical data infrastructure
Valuation Context: $127 million post-money, according to Crunchbase News.

Hang Ten Systems raises $32 million in funding to help large enterprises rebuild around AI

Hang Ten Systems, a Palo Alto-based enterprise AI services startup, has raised $32 million in seed funding to grow its team and expand its work with global enterprises. Hang Ten Systems is the most interpretive deal in today’s set. On paper, it is an enterprise AI services company. In practice, it is a bet on Vishal Sikka’s ability to turn enterprise transformation into a repeatable AI-native operating model. The company says it helps large enterprises build, change, and run business software faster and more cheaply via agentic code generation, reusable skills libraries, and domain depth in areas such as finance, HR, and product development.

That could sound uncomfortably close to consulting if you read it skeptically — and investors should read it skeptically. But the backers matter: Mayfield led the seed round, Aramco Ventures joined strategically, and Sikka brings a track record from SAP and Infosys that few startup founders can match. Hang Ten also says it is already working with companies, including Siemens Gamesa Renewable Energy and Fresenius. Those references suggest that investors are not funding a pitch deck; they are funding an attempt to productize enterprise transformation around AI.

Why include it among the day’s most important rounds? Because it reflects a growing belief that incumbents will not adopt AI through software purchases alone. Many will need a hybrid of software, domain expertise, and execution capacity. If Hang Ten can convert that into a scalable product layer, it could be a category creator. If not, it risks looking like a well-capitalized services firm in startup clothing. That tension is exactly what makes the round worth watching.

Funding Details
Startup: Hang Ten Systems
Investors: Mayfield, Aramco Ventures, and angel investors
Amount Raised: $32 million
Total Raised: $32 million disclosed
Funding Stage: Seed
Funding Date: June 24, 2026
Headquarters: Palo Alto, California, U.S.
Sector: Enterprise AI/transformation software and services.

Attention raises $30 million in funding to move sales AI from note-taking to action-taking

Attention’s pitch gets at one of the clearest shifts in enterprise AI: the market is tiring of tools that observe work without changing outcomes. The company positions itself as an operating layer for revenue teams, not just a system that records calls and summarizes them. Its agents draft and send follow-ups, update CRMs, and run the next play, enabling the company to tie results back to actions its software actually took. That closed loop is a much stronger wedge than another meeting notes product.

The company says it now runs more than 20 million agent actions a month, serves more than 500 customers, has grown ARR 4x year over year, and has increased average contract value 10x over two years. RTP Global led the Series B, with participation from Aglaé Ventures, Eniac, Alven, Linea Ventures, and angels from Attention’s own customer base. Those metrics, especially the move upmarket, explain why investors are willing to keep backing the company as the AI sales market gets more crowded.

This round matters because it suggests AI revenue software is leaving the analysis phase and entering the execution phase. The next wave will not win by producing better summaries of what sales teams already did. It will win by making the next recommendation, carrying it out, and proving the result. Based on Attention’s previously announced $14 million Series A in 2024 and today’s $30 million Series B, the company has disclosed at least $44 million in total funding to date.

Funding Details
Startup: Attention
Investors: RTP Global, Aglaé Ventures, Eniac, Alven, Linea Ventures, and customer angels
Amount Raised: $30 million
Total Raised: At least $44 million disclosed
Funding Stage: Series B
Funding Date: June 24, 2026
Headquarters: New York, New York, U.S.
Sector: Revenue AI/enterprise software.

Runlayer raises $30 million in funding to govern the enterprise agent workforce

Runlayer is a more foundational enterprise AI bet than Attention. Instead of selling a workflow-specific application, it is trying to become the control layer for how employees and AI agents connect to tools, data, and policies across a company. That may sound abstract, but the need is concrete: once teams start using multiple AI clients, MCP servers, custom agents, and embedded agent platforms, security and IT want one place to see what is happening, what data is being touched, and how much it is costing.

Felicis led the Series A, Khosla Ventures participated, and Fortune reported that Khosla wanted “every available dollar” of the round. Runlayer says the financing brings total funding to $42 million and that customers already include Instacart, Gusto, Decagon, Opendoor, dbt Labs, AngelList, Lemonade, and Fortune 500 companies. Those customer logos are doing a lot of work here: they imply the problem is no longer theoretical. Enterprises are already dealing with shadow AI, permissioning, and rogue-agent risk.

Strategically, Runlayer is one of the day’s most revealing deals because it sits in the narrow but valuable band between enablement and control. Founders often pitch AI governance as defensive software. Investors increasingly see it as spending that enables adoption. Fortune also noted that competitors such as Wiz, Palo Alto Networks, and Okta are building agent-governance capabilities, which will quickly test Runlayer’s argument for independence — a neutral cross-provider layer.

Funding Details
Startup: Runlayer
Investors: Felicis, Khosla Ventures
Amount Raised: $30 million
Total Raised: $42 million
Funding Stage: Series A
Funding Date: June 24, 2026
Headquarters: New York, New York, U.S.
Sector: Enterprise AI governance/agent infrastructure.

Coval raises $28 million in funding to build the testing layer for voice AI

Coval is not building voice agents. It is building the infrastructure that makes voice agents safe to deploy at scale. That distinction is why the round deserves attention. In every AI wave, once adoption moves from demos to production, evaluation, monitoring, and quality assurance become standalone companies. Coval says it is the simulation, observability, labeling, and review platform for voice AI and chat agents, with customers including Zoom, Deepgram, and more than 60 other organizations.

The $28 million Series A funding was led by Norwest, with Base10 Partners, Twilio Ventures, and Y Combinator participating. The company says it has raised $31 million in total since launching in 2024 and has run tens of millions of evaluations. That is the sort of data point investors look for in picks-and-shovels companies: repeated usage embedded in customer workflows. Brooke Hopkins’ Waymo background also matters because it gives Coval a credible analogy — bringing simulation discipline from autonomous vehicles into autonomous voice systems.

The wider implication is that voice AI is maturing fast enough to create a spending layer around reliability. Coval cites more than $7 billion invested in voice AI in the first quarter of 2026 alone. When customer-facing AI systems touch call centers, healthcare, banking, and sales, buyers stop asking whether the model sounds good and start asking whether the system can be trusted. That shift usually creates durable software budgets.

Funding Details
Startup: Coval
Investors: Norwest, Base10 Partners, Twilio Ventures, Y Combinator, Swift Ventures
Amount Raised: $28 million
Total Raised: $31 million
Funding Stage: Series A
Funding Date: June 24, 2026
Headquarters: San Francisco, California, U.S.
Sector: Voice AI testing/evaluation infrastructure.

Caplight raises $16 million in funding to build better rails for private markets

Caplight was the day’s non-AI software financing, and it still says a lot about where startup capital is headed. The company provides pricing data, secondary-market intelligence, and transaction infrastructure for private-company shares — an area that becomes more valuable as companies stay private longer and investors struggle with price discovery. In simpler terms, Caplight is building market plumbing for a venture economy that has outgrown spreadsheets, broker relationships, and stale marks.

BlackRock and Fin Capital led the Series A, LEAP Global Partners co-led, and UBS Investment Bank joined as a strategic investor. That investor mix is the real headline. When BlackRock and UBS write checks to a secondary-data startup, they are not buying a speculative story. They are buying infrastructure that could fit into how institutions source information, price exposure, and access private markets. Caplight says its customers collectively manage over $52 trillion in assets, and BlackRock framed the investment in response to rising client demand for transparency and more efficient secondary-market infrastructure.

This round also connects cleanly to the macro backdrop. PitchBook/NVCA says liquidity remains tight across much of venture, and Caplight is pitching itself directly into that friction. Based on Caplight’s previously disclosed $10 million in funding and the new $16 million Series A, the company has disclosed about $26 million in total funding to date.

Funding Details
Startup: Caplight
Investors: BlackRock, Fin Capital, LEAP Global Partners, UBS Investment Bank, DB1 Ventures, Better Tomorrow Ventures, Clocktower Ventures, Dash Fund
Amount Raised: $16 million
Total Raised: About $26 million disclosed
Funding Stage: Series A
Funding Date: June 24, 2026
Headquarters: San Francisco, California, U.S.
Sector: Private-market data and transaction infrastructure.

Seltz raises $12.5 million in funding to build search infrastructure for AI agents

Seltz is a good example of the market’s shift toward deeper infrastructure bets. The company argues that traditional search was designed for humans clicking links, while AI agents need fast, machine-readable, citation-friendly retrieval from the live web. Fortune reported that Seltz has built its own crawler, index, retrieval models, and ranking stack rather than wrapping Google, Bing, or Brave. That is an expensive path, but it also creates the kind of control and margin profile investors want if agents become heavy, high-frequency search users.

Speedinvest and B Capital led the seed round, with participation from Italian Founders Fund, United Ventures, and Bain & Company’s venture arm, Future Back Ventures. Fortune noted that Seltz’s system crawls hundreds of millions of pages a day and returns results in under 200 milliseconds, while the company site positions the product as a “Web Knowledge API” optimized for AI systems rather than human browsing. That framing matters because search is quietly becoming an infrastructure layer again. If agents become the main interface to information retrieval, the search layer becomes strategic, not just convenient.

Investors are clearly aware that this is a crowded field; Fortune pointed to better-funded competitors such as Parallel and Exa. But the Seltz bet is that control of the full search stack can become a moat when customers prioritize latency, citation quality, ownership, and enterprise deployment over just developer convenience. For founders, that is a reminder that the next infrastructure winners may be built where the AI stack is being “re-opened” by agent behavior.

Funding Details
Startup: Seltz
Investors: Speedinvest, B Capital, Italian Founders Fund, United Ventures, Future Back Ventures
Amount Raised: $12.5 million
Total Raised: $12.5 million disclosed
Funding Stage: Seed
Funding Date: June 24, 2026
Headquarters: San Francisco, California, U.S.
Sector: AI search infrastructure/retrieval.

TensorX raises €8 million in funding to build sovereign AI inference capacity in Europe

TensorX is the smallest round on this list by headline size, but one of the most strategically meaningful. The company launched with €8 million committed to Nvidia Blackwell GPUs, including B300 chips, and is pitching itself as a sovereign AI inference platform for Europe’s regulated industries. It runs on dedicated hardware in Dublin and Helsinki, promises zero data retention, and says it is already generating revenue from finance, healthcare, law, and developer-channel demand.

This is exactly the kind of deal that looks modest if you rank rounds by raw dollars and much bigger if you rank them by geopolitical and market significance. Europe’s AI problem is no longer only about model talent. It is also about who controls compute, where data sits, and whether enterprises can access local inference capacity that complies with GDPR and the EU AI Act. TensorX says it is in talks about a further financing facility and plans capacity across Ireland, the U.K., Germany, France, and the Nordics.

Darius Cubed Ventures is backing the launch, and the company says its current €8 million investment is only the opening move, with plans to deploy up to €100 million in Blackwell GPUs. That combination of sovereign-compute positioning and early customer traction is why the round belongs in a premium daily roundup even though larger consumer or fintech financings were available. Investors are increasingly willing to fund “small” rounds that sit close to strategic bottlenecks.

Funding Details
Startup: TensorX
Investors: Darius Cubed Ventures
Amount Raised: €8 million
Total Raised: €8 million disclosed
Funding Stage: Seed
Funding Date: June 24, 2026
Headquarters: Dublin, Ireland
Sector: Sovereign AI infrastructure/inference compute.

What Today’s Funding Activity Reveals

The first pattern is that application AI is no longer being funded as generic “AI software.” Investors are backing very specific operating layers: patient-access automation, financial decision orchestration, sales action engines, agent governance, voice-agent evaluation, and web retrieval for machines. That is a very different market from the one that dominated 2023 and much of 2024, when companies could still raise around broad promises to “apply AI” to work. Today, almost every company on this list is tied to a defined operational system and a measurable buyer pain point.

The second pattern is the rise of the control layer. Runlayer governs agents. Coval validates voice systems. Taktile inserts humans, rules, and context into automated decisions. Assort and xCures both lean on proprietary data and structured workflow logic rather than raw model access alone. These are all answers to the same market problem: enterprises are willing to deploy AI more aggressively, but only if they can trust the output, trace the actions, and fit the system inside industry constraints.

The third pattern is investor crossover. Today’s rounds were not funded only by software VCs. Goldman Sachs, BlackRock, UBS, and Aramco Ventures joined the flow, while Menlo, Khosla, Norwest, and RTP Global stayed active. That matters because it suggests the market increasingly sees these startups as part software, part infrastructure, and part strategic distribution. In concentrated venture markets, crossover investors tend to follow categories that feel inevitable, not experimental.

The fourth pattern is geographic. The U.S. still dominates the daily tape, consistent with Crunchbase’s finding that American companies have captured nearly 80% of global startup financing so far in 2026. But Europe’s presence is not absent; it is selective. Taktile shows Europe can still produce globally relevant enterprise AI winners, and TensorX shows the strongest European story may now be sovereignty rather than consumer scale. That lines up with Crunchbase’s finding that AI now accounts for more than half of European venture funding and with broader state-backed sovereignty pushes such as France’s Tibi program.

Comparative Funding Table

Strategic Takeaways for Founders and Investors

Founders should notice what investors reward today: not general-purpose AI wrappers, but systems tied to measurable operating outcomes. The companies that stood out had one or more of the following: proprietary workflow data, direct attachment to regulated decisions, visible usage at production scale, or a role in controlling AI behavior after deployment. That is a useful filter for fundraising in the second half of 2026. If your story still depends mainly on model novelty, you are selling into the weakest part of this market. If your story is about measurable throughput, compliance, latency, cost control, or domain-specific memory, investor attention is much easier to earn.

Investors, meanwhile, are showing a preference for defensibility below the user interface. Search stack ownership at Seltz, sovereign compute at TensorX, decision logic in Taktile, evaluation in Coval, agent governance in Runlayer, and structured clinical data in xCures all point in the same direction: the premium is shifting toward companies that own a hard layer of the workflow or infrastructure. That is also the best hedge against commoditization as access to foundation models becomes cheaper and more standardized.

There is also a timing signal here. PitchBook/NVCA’s data suggests venture remains highly concentrated and liquidity is still constrained, which means raising large rounds is still easier for companies that can claim category leadership or strategic importance. That makes this a good time to raise if you can prove you’re within a mission-critical budget. It is a much harder time if you are still selling an optional productivity feature.

Finally, today’s mix should remind founders that non-U.S. companies do not need to copy Silicon Valley’s exact playbook to get funded. TensorX’s sovereignty story works because it is native to Europe’s needs. Taktile wins by being deeply tuned to financial institutions rather than chasing broad AI abstractions. The best founders in this cycle are not simply “building with AI.” They are building for the constraints, economics, and power structures of specific buyers.

Conclusion

If there was a single message in today’s funding flow, it was this: venture investors are no longer paying mainly for AI potential. They are paying for AI systems that can survive contact with real organizations. The winners in this 12-hour window were not consumer novelty plays or vague horizontal tools. They were companies embedded in healthcare access, financial risk, sales execution, voice reliability, search infrastructure, private-market rails, and sovereign compute.

That is where the startup market appears to be heading next. The next great venture outcomes may still come from AI, but the capital is moving from fascination to implementation. In plain English, investors want to fund the software and infrastructure that companies cannot operate without once AI becomes part of everyday work. Today’s rounds suggest that the phase is no longer approaching. It has already started.