Storing & Retrieving Knowledge

While the AI learns automatically, you can also explicitly teach it facts. Simply state the information directly — "Our standard hourly rate is $150" or "The Johnson account is managed by Lisa on our team." You can also tell it to remember something — "Remember that ClientCo prefers to be invoiced on the 1st of each month." The AI will confirm it has stored the information and categorize it appropriately. This is the fastest way to populate your knowledge base with critical business information.

Ask your AI employee anything it has learned and it will search its knowledge base to find the answer. You can ask direct questions like "what is ClientCo's billing preference?" or broader ones like "what do I know about the Johnson account?" The AI uses semantic search to find relevant knowledge entries, meaning you do not need to use the exact words you used when storing the information — it understands meaning and context.

Information changes, and your AI employee handles updates gracefully. If you tell it something that contradicts existing knowledge, it will flag the contradiction and ask whether to update the record. You can also explicitly ask it to update a fact — "update ACME's main contact to be Jennifer instead of Sarah." The AI keeps a history of changes so you can see when and why information was updated.

During onboarding, you can import large amounts of knowledge at once by sharing documents, spreadsheets, or lists. Your setup team can also bulk-import contact lists, pricing tables, and other structured data. After initial onboarding, you can continue to add knowledge in bulk by sending documents to your AI employee for ingestion. The AI processes and categorizes the information automatically.