Cantina launches with $8M to take security work from triage to verified fix
UPDATED 09:00 EDT / JULY 30 2026
Agentic security startup Cantina formally launched and said it has raised 16.5 million.
The platform runs AI agents across a customer’s environment covering discovery, triage, remediation and verification that a fix held. Co-founder/CEO Hari Mulackal: AI has “dramatically accelerated the pace of attack”; defender workflows were built for a pre-AI world.
Product
At the center is a “security memory layer”—a running model of the customer environment mapping identities, servers, repositories, cloud permissions, databases, ownership, dependencies, recent changes, and prior investigations. Customers can run stock agents, build their own, or pull community agents from other Cantina customers. Investigations feed back into the model. Platform confirms fixes and keeps auditable evidence.
Cites 2026 Verizon DBIR: vulnerability exploitation overtook stolen credentials as leading breach route; share of critical known-exploited vulnerabilities fully remediated fell to 26% from 38%.
Targets healthcare, fintech and other regulated industries, mid-market to Fortune 500. Says it holds trusted access with OpenAI Group PBC and participates in Anthropic’s Cyber Verification Program.
Customer quote (Trend Health Partners CISO Matt Mock): first platform his team used that carries work from finding an issue through driving the fix.