Knowledge Graph: Guided Onboarding

Every team has its own definition of a qualified lead, its own reporting expectations, campaign guardrails and approval rules. That context decides whether an AI answer is actually useful - and whether the same work can be repeated. The Knowledge Graph now builds it with you, through a conversation.

What's new

• Start from a real use case - define your KPIs, analyze audiences and funnels, or explain why numbers disagree.
• The Builder finds the gaps - it reviews the knowledge you already have and asks about the details that are missing.
• Every proposal argues its case - each one arrives as a card with the evidence behind it and an explanation of why it is asking.
• You decide, one by one - confirm, correct, decline or skip. Only approved knowledge is saved, and the agent reuses it from then on.
• Pause whenever you like - the review stays in the original chat and in Welcome's Suggestions. Once you apply your choices, the agent writes them in the background.

Set your context once - definitions, guardrails, reporting rules - and it carries into future reports, funnel analysis and recommendations: less repeated briefing, more consistent output, and a clearer path from one good conversation to a workflow you can run again.

Check the documentation →