Skip to content

How Personalisation Works

What it does

Every time you chat in AI Training, the AI analyses your messages and extracts meaningful facts, preferences, goals, habits, interests, working style, and more. These are stored as structured memories in Warp's knowledge of me and used by warpSpeed to make every AI interaction more relevant to you specifically.

How to use it

  1. Train regularly: open AI Training and have a conversation about yourself. The AI extracts the meaningful signal from your conversation and categorises it automatically. Examples of what gets saved:
  2. "Prefers concise responses without unnecessary detail" (preference)
  3. "Long-term goal is to grow a product team to 10 people by end of 2025" (goal)
  4. "Likes hiking and outdoor activities on weekends" (interest)

  5. "Works best in short focused sprints rather than long deep-work sessions" (working style)

  6. Let it affect warpSpeed responses: what Warp knows about you is passed as context when warpSpeed generates responses across these features:

  7. Warp: responses are tailored to your communication preferences and context.
  8. Email AI: tone suggestions and draft generation take your communication style into account.
  9. Task AI: time estimates and description generation factor in how you describe your work.

  10. Calendar AI: event description generation uses your professional context.

  11. Keep your profile current: as your circumstances change, return to AI Training and share updates in a new conversation. The AI picks up profile changes immediately.

  12. Edit or delete outdated memories directly in Warp's knowledge of me if something is no longer accurate. One-off questions, filler words, and conversational pleasantries are not saved, only meaningful learnable facts about you.

⚠️ Verification needed: Confirm which specific downstream AI features actively consume the personalisation profile data from the API, and whether the profile is passed on every request or only when relevant context is detected. Flag for engineering review.

Tips & shortcuts

  • The more specific and honest you are during AI Training sessions, the more precisely warpSpeed can tailor responses, vague generalities produce vague personalisation.
  • You don't need to repeat yourself across sessions; the profile accumulates over time, so each conversation adds to what warpSpeed already knows.