Empress

What Empress is

An observability layer for AI agents. It records what each agent did, why it did it and what came out, so an automated action can be explained afterwards.

The point is evidence: when something goes wrong, you can trace it.

What you can do with it

• Log every action an agent takes
• Record the reasoning behind a decision
• Trace the result of an automated step
• Produce logs suitable for audit
• Investigate incidents after the fact
• Monitor agent behaviour over time

Who it is for

• Teams running agents in production
• Compliance and risk functions
• Engineering teams needing traceability
• Organisations with audit obligations

What to watch out for

• Logging agent reasoning means storing prompts and outputs, which may contain sensitive data
• Retention periods need setting to match your audit requirements
• Observability is not control: it records actions rather than preventing them
• Confirm how logs are secured and who can read them