How MindFort is Securing the World's Software in the AI Era

MindFort, building the security agent company, led by Brandon Veiseh and Akul Gupta, has raised a $3M+ seed round, led by Soma Capital, with participation from Y Combinator, 468 Capital, CRV, Sandwith Ventures, and Blast.

The raise comes as Anthropic showcases AI systems capable of uncovering thousands of hidden vulnerabilities, highlighting the limits of traditional security approaches.

MindFort addresses this shift with an autonomous security engineer built for continuous, real-time protection.

“Software is being written faster than it can be secured, and existing tools weren’t built for that pace,” said Brandon Veiseh, co-founder and CEO of MindFort.

MindFort’s agents run autonomous pentests on every CI/CD push or schedule, chaining vulnerabilities into validated exploits and delivering fixes as pull requests via integrations like GitHub, Linear, and Jira.

Since launch, the platform has onboarded customers in minutes and delivered results in hours, outperforming legacy security tools and even human pentesters, with low false positives and rapid remediation times.

Finding vulnerabilities is only part of the challenge, as many issues span cloud configuration, infrastructure, identity policies, network rules, and third-party integrations, not just application code.

MindFort focuses on full-stack remediation, extending its agents to identify and fix issues across all layers, not just through code changes.

Powering the platform is HillClimb, MindFort’s recursive learning system that builds a knowledge graph of each target and compounds insights across engagements.

In parallel, MF-1, a purpose-built model for offensive security, aims to bring frontier AI capabilities to security workflows, improving performance while significantly reducing cost.

The seed funding will support scaling capacity to meet rising demand, expanding the applied AI research team behind MF-1, and extending platform coverage across web, API, network, and cloud environments.