Led · Seed · Jun 2026

Baz develops agentic coding products, including AI Code Review and the newly announced Baz Planner, which intercepts planning-stage changes to detect, root-cause, patch and validate vulnerabilities and bugs before code is authored. Its flagship AI Code Review provides purpose-built agents—Spec Reviewer, Advanced Security Agent, SRE Agent and Fixer Agent—that validate requirements, reason across auth/network boundaries, correlate changes with production telemetry, and automatically apply tested fixes. Baz reports more than 100 leading AI, infrastructure and cybersecurity customers since launching last year and says its AI Code Review is ranked #1 on the precision-weighted Code Review Bench. Customers have reported up to a 65% reduction in downstream rework measured by revert and hotfix PR frequency after adopting Baz Planner. The company’s founding team previously scaled Palo Alto Networks’ Cloud AppSec business and is positioning Baz as a governance and security layer for AI-generated code.

Participated · Seed · Mar 2026

Surf AI provides an agentic platform that reads signals from identity, cloud, security, data, HR, and IT tools to build a living context graph linking assets, owners, permissions, and dependencies. The platform prioritizes risk by real business impact and coordinates remediation through goal-oriented workflows, with specialized AI agents executing under human oversight while preserving context as work progresses. Early deployments have produced measurable results, including nearly $1 million recovered in excess SaaS license spend, thousands of dormant and orphaned accounts disabled, significant certificate risks remediated, and identity enforcement automated at enterprise scale. Founded in 2024, the company is onboarding enterprise technology clients, including global organizations and Fortune 500 companies. Surf AI plans to use its new funding to invest further in product development, expand its team, and scale adoption among enterprises, and it is hiring engineers and security leaders in the U.S. and Israel.

Participated · Series B · Feb 2026

Reco builds an AI SaaS security platform whose proprietary AI Agents continuously monitor enterprise SaaS environments such as Salesforce, ChatGPT, and Microsoft Copilot. The software surfaces real-time insights into permissions, data movement, and unsanctioned “shadow AI” tools, helping security teams mitigate risk. Its AppFactory technology underpins more than 215 out-of-the-box integrations, allowing rapid onboarding of additional cloud applications. Following a 400% increase in annual recurring revenue during 2025, the company now counts multiple Fortune 500 firms and global enterprises in healthcare, financial services, and manufacturing—among them Waste Management—as customers. Reco plans to deploy fresh capital to expand engineering, product, and go-to-market teams, aiming to accelerate platform development and sales reach. Including the newly announced round, the New York City-based startup has raised a cumulative $85 million to date.

Led · Seed · Jan 2026

Topos Bio is building a computational-experimental platform that models intrinsically disordered proteins (IDPs) as dynamic ensembles rather than fixed structures, uncovering transient binding pockets previously considered ‘undruggable’. Its Topos-1 foundation model outperformed leading protein models such as AlphaFold in internal benchmarks and feeds a generative chemistry engine that designs small molecules aimed at modulating these elusive targets. The system is tightly integrated with an in-house wet lab for rapid experimental validation and iterative model improvement. Initial discovery programs concentrate on neurodegenerative diseases, while pipeline expansion into oncology and metabolic disorders is planned. Strategic validation is underway through a collaboration with Gladstone Institutes focused on alpha-synuclein in Parkinson’s disease. The company was founded by UC Berkeley and Stanford researchers with expertise in physics, AI and neuroscience, and it has raised $10.5 million in seed funding to advance its platform.

Led · Series A · Nov 2025

Humanix focuses on protecting the “human layer” of enterprises by detecting manipulation tactics in live conversations across phone, chat, and ticketing systems. Its platform combines AI-powered natural-language analysis with behavioral research to identify social-engineering attempts as they happen, then supplies automated guidance so staff can intervene safely. The system also records proof that proper procedures were followed, reducing compliance exposure and liability for customers and technology partners. Current customers already include leaders in finance and hospitality, underscoring early market traction. By shifting protection from after-the-fact training to real-time intervention, Humanix aims to transform employees from vulnerabilities into assets. The company positions itself to reduce fraud losses that now exceed $100 million annually from help-desk breaches alone. Looking ahead, Humanix plans to extend coverage across every customer-facing channel and to deepen its AI models as attackers adopt more sophisticated tactics.