Launches
Kindo's Deep Hat Sift wants to cut AI bills by skipping unnecessary agent runs
Source-backed 3 sources
Kindo's self-hosted event selection service learns a target agent's verdicts and withholds runs only after human threshold approval.
What happened lately
Deep Hat Sift is an event selection model delivered as a self-hosted service for Kindo agents. It scores alerts, tickets and logs in shadow mode, learns from the target agent’s verdicts, and begins withholding events only after human approval. Kindo describes local CPU scoring, a Kindo API key for integration, and model calls routed through the platform. The current release uses a single-host Docker Compose stack and local accounts without SSO or MFA. Its measured misses are relative to agent judgments, not independently established incident ground truth.