Participated · Series C · Aug 2026

CodeRabbit offers an automated platform that uses AI agents to review freshly generated code in real time, catching syntax errors, logic bugs and security vulnerabilities as code is created. The company has expanded its offering with Agentic Change Management, which includes CodeRabbit Triage for scoring and routing pull requests, Change Stack for visualizing a change's impact on system behavior and dependencies, and CodeRabbit Security for full-repository scanning and continuous monitoring after code ships. Its tools validate incoming pull requests using repository-wide context and sandboxed test environments, and can generate recommended fixes that re-enter the pull request workflow. CodeRabbit has positioned the product to help teams prioritize and govern a rising volume of code changes originating from developers, nontechnical staff and automated agents. The company announced a $143 million Series C to accelerate these product capabilities and scale adoption. Prior investors such as CRV, Scale Venture Partners, Flex Capital and Pelion Venture Partners participated in the latest round.

Participated · Series B · Jun 2026

Founded in 2023 and based in San Francisco by former Meta AI researchers Anand Kannappan and Rebecca Qian, Patronus AI creates simulated digital environments that replicate websites and internal systems to evaluate AI agent behavior. The company uses these digital world models to stress-test agents with reinforcement learning, enabling iterative rewards for task completion and penalties for errors. Patronus currently provides simulations for software engineering and finance, and reports that demand from frontier AI labs and startups is nearly insatiable. Its revenue has grown 15-fold over the past year, contributing to strong investor interest. Patronus positions itself primarily against internal evaluation teams at AI labs and differentiates from human-data firms by evaluating agent behavior without human involvement. The company aims to build environments capable of running long-duration agent operations across a range of verifiable and, eventually, more complex scenarios.

Participated · Series A · Jun 2026

Aryon Security offers a Cloud Security Enforcement Platform that shifts cloud security from reactive detection and remediation to preventive enforcement before risks reach production. Since emerging from stealth in 2025, the company says its platform can avoid up to 95% of traditional CSPM alerts and has gained traction with mid-to-large enterprises across regulated industries such as healthcare, banking, insurance, telecommunications, shipping, and industrials. Founders Ron Arbel (CEO), Ariel Litmanovich (CTO), and Yair Ladizhensky (CPO) are veterans of the IDF’s Matzov cybersecurity unit and Forbes 30 Under 30 honorees who worked on securing Project Nimbus. The company reports three primary use cases: secure cloud migration and M&A, preventing new gaps in mature complex environments, and operationalizing prevention at scale with granular, production-safe enforcement and exception management. Aryon also stated a long-term vision to extend preventive enforcement beyond cloud infrastructure to broader enterprise environments as organizations adopt more AI, SaaS, and distributed technologies. Financially, Aryon has raised $29 million in Series A and $38 million in total capital to date.

Participated · Series A · May 2026

MokN develops technology to detect and recover compromised identities, positioning itself in the emerging category it calls Active Identity Recovery. Its first product, Baits, deploys realistic decoy access points to reveal attackers using stolen credentials. The company says it currently protects more than one million users across large enterprises and mid-sized organisations. MokN plans upcoming products to secure customer accounts, stolen browser cookies and compromised sessions. Financially, it completed a €2.6 million seed round in October 2025 and has now raised $15 million in a Series A. The firm intends to invest the new capital in R&D, expand its product suite, and grow operations in France, the United States, and the United Kingdom.

Participated · Series B · May 2026

Dust builds a shared collaboration surface and orchestration platform where teams and AI agents work together with shared projects, context, conversations, to-dos, notifications, and cloud compute for processing files and generating documents. An intelligence layer integrates more than 100 data sources and common tools so agents can act on company context, while built-in memory and reinforcement loops allow agents to learn and improve over time. The platform includes enterprise governance features—granular permissions, cost and usage monitoring, full audit trails, and agent analytics—and is SOC 2 Type II certified and GDPR compliant with EU and US data residency; Dust also contractually guarantees that customer data is not used to train models. Dust reports usage by more than 3,000 organizations, over 300,000 deployed agents, 70% weekly active usage, and zero churn in 2025. The company was founded by Gabriel Hubert and Stanislas Polu, who previously worked together at Stripe and in AI research, and has raised over $60 million to date. Going forward, Dust plans to focus on automation-driven agent improvement, deeper human-agent collaboration primitives, and enterprise-grade governance and orchestration infrastructure.