FriskAI raises $3.6M to watch what AI agents do in production
What's the deal? FriskAIDealroom has a profile for this one. Try Dealroom → emerged from stealth with $3.6 million in pre-seed funding to track what AI agents do once deployed. MaC Venture CapitalDealroom has a profile for this one. Try Dealroom → led the round, joined by Wischoff VenturesDealroom has a profile for this one. Try Dealroom → and NEA partner Rick YangDealroom has a profile for this one. Try Dealroom →. The Los Angeles startup targets the healthcare, insurance, and finance sectors.
How it works: FriskAI sits between an agent and the tools it touches, logging every action so security and compliance teams can see what happened, when, and whether it fell outside expected behaviour.
Why now? Enterprises have spent the past year handing AI agents access to their databases, internal tools, and customer systems. Almost none can say with confidence what those agents did with it. Regulated industries are moving agents out of pilots into production and finding their monitoring tools were built for software that follows a script — not one that decides things on its own.
What's the endgame? Founder and chief executive Neel Palrecha, who previously shipped production systems at Apple, HeadspaceDealroom has a profile for this one. Try Dealroom →, Snapdocs, and PlayVSDealroom has a profile for this one. Try Dealroom →, wants to make agents visible. "Companies can monitor servers, applications, and employees, but they lack runtime intelligence for AI agents," he said. The funding will expand engineering and go-to-market teams.
By the numbers: The round lands in roughly the 79th percentile for deal size — a healthy pre-seed. But FriskAI is a small player in a field that has raised over $175 million in 2026. Rivals include Geordie AIDealroom has a profile for this one. Try Dealroom → ($30 million Series A for agent governance), NeuralTrustDealroom has a profile for this one. Try Dealroom → ($20 million), Oasis Security ($120 million for machine-identity management), and InfiniteWatch ($4 million for agent observability).
What could go wrong? Most rivals solve something adjacent — identity, endpoint defence, customer-facing monitoring. FriskAI's narrower bet is that a plain, readable record of what an agent did is what enterprises want first. Whether $3.6 million is enough to build that trust before Geordie or Oasis lock down the same customers is another question.
The signal: The agentic AI security market is projected to grow from $1.65 billion in 2026 to $13.52 billion by 2032, according to MarketsandMarkets. MaC managing partner Marlon Nichols compares the moment to earlier infrastructure booms. "Every technology cycle mints a small group of companies that sell what everyone else ends up depending on, like AWS in the Cloud era and Cisco during the internet boom," he said. The catch: "none of that happens until enterprises can put real weight behind them and trust them with decisions that matter."
Read more: Tech Funding News
Image credit: FriskAI
