RFO.Reason for Outage
by Christy Steyn
Post-mortems

Every failure,
written up.

Browse by what you are building.

01 / Architecture

Retrieval, ingestion, agents.

The design decisions and the boundaries between components.

02 / Failure analysis

Follow the failure.

What broke, what the traces showed, and what reduced the risk.

03 / Observability

Know what it did.

Tracing, golden evals, and review gates around AI systems.