Skip to main content
Ankra AI gets better at working with you in two ways: it remembers what you tell it, so you don’t repeat yourself, and it records how your team rates its answers, so you can see where it falls short. This page covers both, and how to review or delete what is kept.

Memory

Ankra AI keeps a small personal memory: what it learned about you (how you like answers delivered) and saved memories (facts worth keeping across conversations). Review both under Ankra AI → Memory (/next/organisation/ai-memory), in the sections What Ankra AI learned about you and Saved memories.
Memory is personal. The page says it: what Ankra AI remembers about you between conversations, and only you can see it. It shapes your own conversations only.

What Ankra AI learned about you

As you chat, Ankra AI picks up how you work - how you like answers written, how much detail you want, your experience, what you know well, the tools you use, how you work and what you often ask about - and applies it to later answers. Each of these is listed, and you can forget any one of them on its own. The Learn from my conversations switch pauses learning: turn it off and Ankra AI keeps what it learned and adds nothing new. The page confirms Learning is paused. Ankra AI keeps what it learned and adds nothing new. Forget everything it learned… starts over. Your saved memories are kept, and learning stays as it is. This cannot be undone.

Saved memories

Memories are short facts saved during conversations - either because you asked Ankra AI to remember something or because it judged a detail worth keeping. A new conversation starts with them. In chat, Ankra AI can also recall them on demand (the save_memory and search_memories MCP tools expose the same store to connected editors). Each memory carries a badge:
  • Chat only - only used when you chat.
  • Trusted - unattended runs (AI tickets, alert fixes, watches) can use this memory too.
Trust this memory on a Chat only memory asks Trust this memory?: you vouch for this exact text, and once trusted it can steer runs nobody watches. Trust only a memory whose text you would want every unattended run to follow. Delete a single memory with its delete button, or all of them with Delete all memories…. New conversations then start without them; what Ankra AI learned about how you work is kept. Deletion cannot be undone.
Open Ankra AI → Memory and use the buttons above.

Scope

Memory is scoped per user, per organisation: what Ankra AI learns about you in one organisation does not follow you into another.

Feedback

Every answer from Ankra AI can be rated Good answer or Bad answer with the buttons under it. Ankra AI → Feedback (/next/organisation/ai-feedback) shows how people rate Ankra AI’s answers, and what they said about them.

Who sees what

  • People who manage Ankra AI, such as owners and admins, see everyone’s ratings in the organisation and can Export as CSV.
  • Everyone else sees their own ratings only.

Summary

Choose the period and the cluster (Every cluster, or one) at the top. Summary shows the number of Ratings, how many were Helpful and Not helpful, how many came With a comment, The latest days with ratings, and the Clusters rated most.

Every rating

Every rating lists the ratings newest first. Filter them to Every rating, Helpful or Not helpful, or search the comments. Open one to read What was asked, What Ankra AI answered and the Comment. A rating with a short comment (“wrong namespace”, “fixed it first try”) is the most useful signal.

Next

Teach Ankra AI your organisation’s conventions, rather than one person’s, with AI Skills.