Kairos puts risk tooling on display

Kairos is an open-source platform for making fraud and risk decisions as events happen. Its repository describes a system that can return an approve, step-up, review or block decision, with a risk score and reasons. It also outlines workflows, models, and tools for people to review cases. You can use its cloud service or self-host it with Docker, according to the project repository.

The project drew 103 points and 30 comments on Hacker News. That’s an early sign of community interest, not proof that companies are using it. The repository doesn’t provide user details, so there’s no basis here to judge adoption or results in real businesses.

For an operator, the practical question is whether a system like this fits your existing fraud process, and who would manage its rules, models and review queues. Kairos says it can connect with coding agents to build workflows and test model options, but that’s a capability described by the project, not evidence of business outcomes. For operators, it raises the question of how a tool like this would fit alongside other tools and workflows in an AI command centre.