Venture capital & startup funding roundup, July 20, 2026: Accel, a16z, Bessemer, Kleiner Perkins, and NEA

Venture capital did not scatter broadly across the startup market today. It moved with a clear bias toward infrastructure, security, and software that sits in the control plane of AI rather than the presentation layer. The biggest check of the cycle went to CuspAI’s push into AI-driven materials discovery, while other notable rounds clustered around cybersecurity, inference software, payments rails for AI agents, and regulated workflow automation. That mix matters. It suggests investors still want exposure to AI, but they increasingly prefer businesses that shape compute economics, data control, or mission-critical enterprise processes rather than another thin wrapper on top of a frontier model.

There was another signal embedded in the day’s tape: public disclosure was uneven. Several of the most strategically interesting transactions were not classic, price-tagged venture rounds at all, but round extensions or strategic investments from private-equity, corporate, or ecosystem partners. That is often what a more selective market looks like. The money is still there, but it is concentrating around founders with technical leverage, distribution leverage, or both. In a strict 12-hour screen ending at 4:25 PM ET, that meant a smaller fully disclosed pool than the headline appetite for AI might imply.

For founders, the message is not simply “build in AI.” It is “build where AI creates a new bottleneck.” Today’s rounds rewarded companies working on chip readiness, supply-side healthcare logistics, cyber defense for AI-era applications, and systems that let autonomous software move money or decisions safely. Investors are still paying for ambition, but only when ambition is attached to a bottleneck customers will budget for.

The Macro Environment: Capital Crowds Into the AI Control Layer

The strongest theme today was a migration from broad AI enthusiasm into the AI control layer: materials, chips, security, payments, and policy workflows. CuspAI’s $450 million Series B is the clearest expression of that move. Investors are not funding a generic model company there; they are backing a bet that AI can shorten the physics-and-chemistry loop that underpins semiconductors, clean energy, and advanced manufacturing. Infinity’s seed round fits the same pattern from another angle, focusing on the software layer that makes new AI silicon inference-ready. Natural, meanwhile, is tackling the transactional plumbing required for AI agents to spend money in the real world.

Cybersecurity was the second major cluster, and it was not incidental. Neo and Empirical Security both raised on the thesis that AI is changing the shape of enterprise risk: software is becoming more autonomous, vulnerabilities are surfacing faster, and security teams need systems that can see and respond to that shift in real time. Investors have long treated security as a resilient category; what changed today is that the spend case now ties directly to agentic software, application permissions, and exploit prediction. That makes these rounds less cyclical than they might appear on the surface.

A third pattern was the return of milestone-driven financing. Brenus Pharma’s Series A extension and Quorum’s strategic investment both show how capital behaves in a more selective environment. Instead of rewarding raw narrative alone, investors are leaning toward follow-on financings that accelerate companies already demonstrating traction, de-risking, or category leadership. Brenus tied its extension to clinical, regulatory, and business-development progress. Quorum’s backer framed its investment around product expansion and agentic AI capabilities in a market already anchored by thousands of users.

Geography also mattered. The largest check went to the U.K. CuspAI. France showed up through Brenus Pharma. Boston and Chicago contributed cybersecurity and infrastructure-adjacent software. India’s Plazza round underlined that investors still see operationally dense, real-world delivery networks as investable when the problem is urgent, and frequency is high. This was not a one-city market. It was a market rewarding whichever founders could convert technical edge into a priced, defensible system.

The Top Funding Rounds

CuspAI raises $450 million to build AI-driven materials discovery for semiconductors, energy, and climate

CuspAI founders (Credit: CuspAI)

CuspAI’s Series B was the day’s defining transaction because it points to where deep-pocketed investors think the next AI moat will sit: not only in models, but in the physical systems those models help design. The company uses AI for materials discovery, targeting applications across semiconductors, batteries, clean energy, and advanced manufacturing. The round valued the Cambridge startup at $2.6 billion and expands a business that has already assembled a high-profile network of partners and advisers.

Investors care because materials are upstream of several constrained markets at once. If AI can reduce dependence on rare metals in chipmaking, shorten R&D cycles, or improve energy materials, the payoff is not confined to software multiples. It cascades into manufacturing economics, supply-chain resilience, and geopolitical competitiveness. That is why a venture syndicate led by Kleiner Perkins and NEA was joined by Bezos Expeditions, the U.K. government, AMD Ventures, Lux Capital, Glade Brook Capital Partners, and Invest-NL.

What makes this round especially notable is that it pushed CuspAI past $650 million in total raised only two years after launch. That kind of capital intensity would look reckless in a consumer app, but in AI-for-science it reads differently: compute, lab workflows, and industrial partnerships are expensive, and investors appear willing to fund the stack if the upside touches trillion-dollar industries. For founders, the lesson is that “deep tech” only gets this kind of check when it is tied to strategic industrial demand, not abstract scientific promise.

Funding Details
Startup: CuspAI
Investors: Kleiner Perkins, NEA, Bezos Expeditions, the U.K. government, AMD Ventures, Lux Capital, Glade Brook Capital Partners, Invest-NL, and others
Amount Raised: $450 million
Total Raised: More than $650 million
Funding Stage: Series B
Funding Date: July 20, 2026
Headquarters: Cambridge, United Kingdom
Sector: AI for materials discovery / advanced materials / industrial AI

Neo raises $100 million to secure enterprises as AI turns software into autonomous systems

Neo’s $100 million funding from stealth shows how much investors still value founder credibility in cybersecurity, especially when the technical change is as large as AI-enabled software autonomy. The company was founded in Boston by former SentinelOne executives Nick Warner and Shlomi Salem alongside technologist Eran Shirazi, and the pitch is direct: traditional enterprise security tools are poorly matched to a world of AI-enabled applications and agentic systems.

The backing roster matters here. Andreessen Horowitz, Bessemer Venture Partners, Craft Ventures, and Merlin Ventures are not making a small option bet; they are underwriting the idea that identity, permissions, and visibility will need to be rebuilt for AI-era applications. Neo’s platform is designed to let security teams see, review, and control AI-enabled software before data access or automated actions create new operational risk. The company has already piloted its technology in finance, energy, and transportation, which suggests buyers are testing these tools before the problem becomes fully mainstream.

Strategically, Neo fits a broader rule of the current market: investors will still write very large early checks when the founding team has category-level pattern recognition and the threat model is changing fast. Security is rarely a discretionary category, and AI is increasing both budget urgency and product complexity. That is why this round feels less like a stealth launch and more like a preemption play.

Funding Details
Startup: Neo
Investors: Andreessen Horowitz, Bessemer Venture Partners, Craft Ventures, Merlin Ventures, and other backers not fully disclosed in the accessible report
Amount Raised: $100 million
Total Raised: $100 million
Funding Stage: Combined seed and Series A financing
Funding Date: July 20, 2026
Headquarters: Boston, United States
Sector: Cybersecurity / AI application security

Natural raises $30 million in funding to build payments infrastructure for AI agents

Natural is attacking a problem that will grow with every credible AI agent workflow: how software acts financially on behalf of a user or company without creating permission chaos, payment friction, or compliance headaches. The company announced a $30 million Series A led by Kirsten Green at Forerunner, and TechCrunch reported that the round brings total funding to $40 million.

Why investors care is straightforward. Agentic commerce is easy to demo and hard to operationalize. Once software begins purchasing software, paying vendors, or handling transactional workflows, the underlying rails become the product. Whoever owns that layer can capture volume, compliance, and embedded distribution far beyond what a thin productivity app can capture. That gives Natural a more durable position than many AI application startups whose differentiation may erode as base models improve.

This round also shows that consumer-oriented investors are not stepping away from AI; they are simply moving toward economically central AI use cases. Payments is one of those cases. Founders should notice the distinction. AI that saves a click may struggle. AI that safely moves a dollar can command much more strategic capital.

Funding Details
Startup: Natural
Investors: Forerunner, led by Kirsten Green; additional participants were not fully disclosed in the accessible public materials reviewed
Amount Raised: $30 million
Total Raised: $40 million
Funding Stage: Series A
Funding Date: July 20, 2026
Headquarters: Not disclosed in the accessible public materials reviewed
Sector: Fintech / payments infrastructure / agentic AI

Empirical Security raises $25 million in funding to help security teams predict exploited threats

Empirical Security’s Series A is another sign that cyber buyers want systems built for the speed and messiness of the AI era, not just prettier dashboards on top of old workflows. Axios reported that the Chicago company raised $25 million in Series A funding led by Brightmind Partners, while follow-on posts from existing backers said the financing brings the company’s total funding to $37 million.

The company’s positioning is notable. Rather than centering on broad AI security rhetoric, Empirical focuses on using AI to help companies predict threats by monitoring exploited vulnerabilities. That is a narrower, more operational claim, and that is likely part of why it attracted capital. Budgets tend to open faster for software that helps teams prioritize concrete exposures than for platforms that merely promise “more intelligence.”

Investor behavior here is just as important as the product. Brightmind led, while prior investors HPA and Costanoa Ventures also participated. That combination suggests continuity capital is still alive when a company has a clear product wedge and measurable urgency. In a crowded cyber market, the winning story is increasingly not “we do AI security,” but “we reduce exploit pressure where teams already feel pain.”

Funding Details
Startup: Empirical Security
Investors: Brightmind Partners, HPA, Costanoa Ventures, and others
Amount Raised: $25 million
Total Raised: $37 million
Funding Stage: Series A
Funding Date: July 20, 2026
Headquarters: Chicago, United States
Sector: Cybersecurity/exposure management / AI-assisted threat prediction

Infinity raises $15 million in funding to make any AI chip inference-ready

Infinity’s seed round stands out because it attacks a market bottleneck that sits between the explosion of AI silicon startups and their ability to win real workloads. The company says it is building the software layer that makes any AI chip inference-ready, and Business Wire said the financing totaled $15 million at a $100 million post-money valuation.

That sounds technical, but the investment case is easy to understand. New chips do not matter if developers cannot deploy on them quickly. Nvidia’s grip on AI is partly about software and ecosystem maturity, not only hardware performance. Infinity is effectively selling time-to-utility: if new silicon vendors can become inference-ready in days instead of months or years, they have a better chance of competing for production demand. That makes Infinity a leverage play on the broader AI hardware wave without requiring it to manufacture chips itself.

The deeper signal is that venture capital is taking the “second-order AI stack” more seriously. The market has already spent heavily on model builders and chip companies. The next money is flowing toward the translators, adapters, and orchestration layers that make that infrastructure usable. Infinity is a clean example of that shift.

Funding Details
Startup: Infinity
Investors: Not disclosed in the accessible public announcement reviewed
Amount Raised: $15 million
Total Raised: $15 million
Funding Stage: Seed
Funding Date: July 20, 2026
Headquarters: San Francisco, United States
Sector: AI infrastructure/inference software/semiconductors

Brenus Pharma raises €11 million in funding to advance cancer immunotherapy and extend its Series A

Brenus Pharma’s financing is a useful reminder that biotech capital has not disappeared; it has become more milestone-sensitive. Business Wire said the Lyon company added €11 million to its Series A, bringing total capital raised since inception to €38 million. The company tied the extension to execution across clinical, regulatory, and business-development milestones around STC-1010, its lead clinical-stage immunotherapy program for gastric and colorectal cancer.

Extensions like this matter because they tell you how investors are underwriting risk. Rather than forcing every company into a fresh narrative reset, backers are willing to add capital when a management team has de-risked the science enough to justify another tranche. That is often healthier than a brand-new round at an inflated mark because it links capital directly to progress.

There is also a geographic signal worth noticing. The company explicitly said it welcomed new European and Asia-Pacific life sciences investors. In an era when biotech capital can look heavily U.S.-centric, that matters. It suggests that clinically grounded assets with a visible regulatory path can still pull cross-border investor attention, even in a selective market.

Funding Details
Startup: Brenus Pharma
Investors: New European and Asia-Pacific life sciences investors; the full investor roster was not fully disclosed in the accessible teaser reviewed
Amount Raised: €11 million
Total Raised: €38 million
Funding Stage: Series A extension
Funding Date: July 20, 2026
Headquarters: Lyon, France
Sector: Biotech/immunotherapy/oncology

Quorum lands an undisclosed strategic investment to expand agentic AI for government affairs

Quorum’s financing was not the day’s largest check, but it may be one of the more revealing. Enlightenment Capital announced a strategic investment in the Washington company, describing Quorum as an AI-powered government affairs platform used by more than 2,000 organizations, including over half of the Fortune 100. The public announcement did not disclose the dollar amount.

Investors are paying attention here because the company sits where policy volatility meets enterprise workflow. Public affairs and government-relations teams are under pressure to process more legislative activity, stakeholder noise, and regulatory change with smaller teams and tighter timeframes. Quorum’s appeal is that it turns what used to be fragmented research, stakeholder management, advocacy, and PAC operations into a single operating system. In a period of heavy policy flux, that looks a lot like infrastructure.

The strategic angle also matters. This was not framed as a trophy growth round. It was framed as capital to accelerate product roadmap execution and expand agentic AI capabilities. That is exactly the kind of investment structure that tends to show up when a company is already embedded in high-value workflows and needs a partner more than a publicity spike.

Funding Details
Startup: Quorum
Investors: Enlightenment Capital
Amount Raised: Undisclosed
Total Raised: Undisclosed
Funding Stage: Strategic investment
Funding Date: July 20, 2026
Headquarters: Washington, D.C., United States
Sector: Government technology/enterprise software / AI for public affairs

Wagmo lands an undisclosed strategic investment to push pet health benefits through credit unions

Wagmo’s announcement shows a very different but still instructive investment pattern: distribution-led strategic capital. MorningStar surfaced a strategic investment announcement involving Wagmo and Curql, framed around bringing modern pet healthcare into the credit union channel. Publicly accessible materials did not disclose the amount, and the teaser available in search did not expose the full timestamp, but the deal appeared among Business Wire’s latest July 20 results. Wagmo’s own company profile describes it as a New York-based pet healthcare company.

Why it matters is less about pet insurance specifically and more about channel strategy. A lot of insurtech and benefits startups still think in terms of direct response or employer sales. This deal points to another route: financial institutions as trusted distribution for adjacent household products. Credit unions want differentiated member benefits. Wagmo wants lower-cost acquisition and embedded trust. That kind of partnership can be more durable than paid acquisition if it works.

From a venture perspective, strategic capital like this tends to appear when the next leg of growth depends on distribution more than raw product iteration. Founders should pay attention. Sometimes the most valuable investor is not the one who pays the highest price, but the one who opens the cheapest, most defensible route to customers.

Funding Details
Startup: Wagmo
Investors: Curql
Amount Raised: Undisclosed
Total Raised: Undisclosed
Funding Stage: Strategic investment
Funding Date: July 20, 2026
Headquarters: New York, United States
Sector: Insurtech / pet healthcare / embedded benefits

HALO X-ray Technologies raises a multimillion investment round to complete regulatory approval for security screening

This financing sits just outside a strict 12-hour ET cutoff, but it belongs in the day’s conversation because it captures a rising pattern in deep tech: insiders funding the next de-risking step rather than waiting for a broad new syndicate. HALO X-ray Technologies announced a multimillion-dollar round led by Agilent, with prior investors UK Innovation Science Seed Fund and the Midland Engine Investment Fund also participating, to support regulatory approval and commercialization of its screening technology.

The technology case is compelling because HALO is not selling generic “AI security.” It is building advanced X-ray diffraction capabilities intended to improve material understanding during security screening, which has direct implications for airports, borders, and critical infrastructure. The fact that existing investors doubled down rather than stepping back is meaningful. In hard-tech security markets, follow-on conviction from technically informed investors often carries more signal than a flashy new lead name.

For founders, the deeper lesson is that regulatory and commercial milestones can be financeable when the buyers are obvious and the transition from lab promise to deployed system is believable. That is one of the few ways physical-security startups can still draw meaningful capital in a software-dominated cycle.

Funding Details
Startup: HALO X-ray Technologies
Investors: Agilent Technologies, UK Innovation Science Seed Fund, Midland Engine Investment Fund
Amount Raised: Multimillion-dollar round
Total Raised: Undisclosed
Funding Stage: Strategic / seed-stage follow-on investment
Funding Date: July 20, 2026 release cycle, announced at 3:11 AM ET
Headquarters: Nottingham, United Kingdom
Sector: Security technology / industrial imaging / deep tech

Plazza raises $15 million in funding to expand rapid medicine delivery infrastructure in India

Plazza is another near-cutoff addition, but strategically it was too relevant to ignore. India’s quick-commerce startup announced a $15 million Series A led by Accel, Elevation Capital, and Nexus Venture Partners to expand its pharmacy network and instant medicine delivery operations. Entrackr timestamped the announcement at 10:15 IST on July 20, which places it outside a strict Eastern Time 12-hour screen, but squarely inside the same global funding cycle.

This round works because it is not a vague wellness bet. It is a logistics and trust bet in a category where reliability matters more than brand storytelling. Consumers do not comparison-shop carefully when a prescription needs to arrive quickly. That makes availability, fill rates, inventory routing, and neighborhood density the real moat. Investors in India have been cautious with pure convenience plays, but health delivery can support stronger repeat behavior and clearer urgency.

Plazza also illustrates how healthcare and commerce are converging outside the U.S. venture script. In many markets, the winning product is not a digital therapeutic or a premium care experience. It is the fastest trustworthy bridge between offline medical supply and online demand. That is a very different, and often more scalable, infrastructure story.

Funding Details
Startup: Plazza
Investors: Accel, Elevation Capital, Nexus Venture Partners
Amount Raised: $15 million
Total Raised: At least $16.4 million including earlier seed capital
Funding Stage: Series A
Funding Date: July 20, 2026 local release cycle
Headquarters: Bengaluru, India
Sector: Health commerce/logistics / digital pharmacy

What Today’s Funding Activity Reveals

The most obvious pattern is capital concentration. Only one truly massive round printed, and it went to a company solving a problem upstream of several industrial bottlenecks. Beneath that, the bulk of activity centered on the AI operating layer: cyber defense, chip software enablement, payments rails, and policy workflows. That is not an accident. Investors appear to be rotating away from businesses whose value depends mostly on model access and toward companies that shape cost, security, or workflow control.

The second pattern is that investor type matters more now. Traditional VC led some of the day’s rounds, but a noticeable share of the most strategically important financings came from private-equity or corporate-adjacent capital. Enlightenment Capital backed Quorum. Agilent led HALO’s round. Curql backed Wagmo. That says a lot about the market. When buyers, channels, or domain experts can underwrite part of the next growth chapter, founders are less dependent on pure financial sponsors. In a tighter capital market, that can be an advantage.

Cybersecurity also keeps acting like a magnet category. Neo and Empirical Security pulled large early capital because AI is not just creating new software categories; it is rewriting the risk model for existing ones. Security vendors that can articulate exactly where AI changes permissions, exploitability, or response speed are finding receptive investors. The implication is that cyber remains one of the clearest ways to monetize AI urgency without relying on end-user novelty.

Finally, this was a geographically mixed tape, but not an evenly distributed one. The U.S. still dominated early-stage software and cyber. Europe showed strength in deep tech, biotech, and government-adjacent software. India’s appearance came through operational health commerce rather than frontier AI. That is a useful reminder that global venture is not converging on one template. Different regions are attracting capital where they already have talent density, regulatory familiarity, or operational edge.

Venture Funding Table

Note: HALO and Plazza are included as near-cutoff additions because the strict 12-hour Eastern Time screen produced a thinner fully disclosed set than the requested top-10 list; both were part of the same July 20 global funding tape and were strategically important enough to include with clear labeling.

Strategic Takeaways for Founders and Investors

Founders should take one hard lesson from today’s activity: capital is chasing bottlenecks, not slogans. If your startup touches infrastructure costs, security posture, compliance workflow, or high-urgency fulfillment, investors can still underwrite large rounds. If your differentiation depends mainly on packaging the same model access as everyone else, the bar is much higher. CuspAI, Infinity, Natural, and Empirical Security each won funding by positioning themselves at a pressure point customers will pay to relieve.

Investors, for their part, are signaling a preference for defensibility that compounds. In today’s set, that defensibility came in several forms: scientific IP and industrial partnerships at CuspAI, founder pedigree and urgency at Neo, ecosystem leverage at Quorum and Wagmo, and milestone-based scientific progress at Brenus. The common thread is not just technology; it is the ability to make replacement painful. That is where pricing power survives as AI capabilities become more widely available.

There is also a timing signal in the structure of these deals. More companies are taking extensions, strategic investments, or concentrated early rounds rather than relying on broad, hype-fueled syndication. That usually means both sides of the market are being more disciplined. Founders should expect investors to ask how the next dollar changes outcome, not just runway. Investors should expect the best companies to use partner selection itself as a product and distribution decision.

The final warning is about commoditization risk. AI applications built on broadly accessible models may still grow, but the market is telling founders that the durable money is moving underneath or around the model layer. Payments, security, chips, science, and workflow infrastructure all benefit from AI while remaining harder to commoditize. That is where the premium multiples are likely to concentrate next.

Conclusion

Today’s funding tape did not show a venture market chasing novelty for its own sake. It showed a market trying to price where AI creates durable scarcity. In some cases that was scarce scientific capability, as with CuspAI. In others it was scarce trust, as with cyber and policy software. In still others it was scarce operational reliability, as with medicine delivery and embedded pet-health distribution.

That is the bigger shift founders and investors should watch. The next phase of startup financing looks less like a race to attach AI to everything and more like a contest to own the systems that make AI safe, deployable, and economically useful. The companies that win capital now are not just promising automation. They are defining who controls the choke points around it.