Agnost AI
Find recurring AI-agent failures in conversations and traces
Observe, evaluate, and govern AI agent activity

Traccia provides a shared control layer for teams operating AI agents. It tracks agents, execution traces, model calls, tool use, latency, and spending across supported frameworks. A registry records ownership and environment, while dashboards help teams identify failures, expensive tasks, or behavior that needs investigation.
The platform also supports versioned prompts, evaluation datasets, scorers, and comparisons between experiments. Policy controls can monitor or block configured actions, model choices, and spending limits, with enforcement coverage depending on the SDK or gateway integration. Teams can annotate sensitive-data findings and export activity evidence for review. These features support an organization's governance process, while policy configuration and compliance judgments remain responsibilities of the team using the system.
Tried it? Share a useful thought or ask a question.
Good conversations start with a hello.
Sign in to comment