Memory

Memory is the AI employee's persistent recollection of past interactions, preferences, decisions, and facts — stored outside the context window and recalled on demand.

The model itself is stateless: it forgets every call. Memory is how an AI employee accumulates a working knowledge of you, your company, and prior decisions. Implementations range from simple key-value stores to vector-indexed conversation logs to structured profile databases. Memory is what turns a session bot into a colleague.

Example

On Tuesday you tell the AI EA you prefer 25-minute meetings. On Friday it books one, defaults to 25 minutes without asking, and notes the decision.

How OpenLabor uses it

OpenLabor employees keep persistent memory per workspace — what they learned about you survives sessions.

Can I see and edit what the AI remembers about me?

You should be able to. Quality memory systems expose the entries and let you correct or delete them.

Related: context-window, rag, ai-employee.

AI Labor Glossary