AI product
Datadog Agent Observability is an AI-agent observability product for evaluating, improving, and tracing agents across development and production. It combines offline experimentation with production observability: teams can evaluate agents in production, build golden datasets from annotated traces, and test changes against those datasets before release. The same Datadog tracer can be used from development through production, with tracing across requests, services, agent reasoning, upstream and downstream dependencies, and end-user experience. It also provides alerting, role-based access control, sensitive-data protection, and HIPAA compliance controls. The product addresses agent quality, cost, safety, and operational behavior in addition to conventional application signals such as latency, errors, traffic, and saturation.
1 use taken from transcripts — each links to the moment in the video.
An observability product for understanding whether an AI application is healthy, running properly, and needs fixes.
1 in the library.