What team memory is (and why your wiki isn’t it)
Your team writes everything down and still can’t find anything. The problem isn’t discipline. It’s that documents freeze while the facts keep moving.
Read postBlog
Notes on shared memory for teams and their AI agents — product updates, engineering write-ups, and how-tos.
We tested our memory against a folder of docs. It tied.
A blind test: structured memory versus an AI agent grepping through files. The score was 90–90, with zero hallucination on either side. Here’s why we’re publishing that anyway.
Read postConnect Claude to your team’s memory in five minutes
No API keys to paste, nothing to install. Add the connector, sign in, and Claude can read and write your memories.
Read postWhy your AI agents keep asking the same questions
Every new chat starts from zero. You explain your stack, your customers, your naming rules — again. The missing piece is shared memory.
Read postWhy a human approves everything before it enters memory
LLM extraction is good. It’s not good enough to write into a memory your whole team trusts. So nothing lands without review.
Read postThe meeting ends. The memory updates.
Agents in Ontonym do the boring part of remembering: they read what happened, draft the updates, and send them to review.
Read postStop managing access by hand
Connect your identity provider and your groups become roles and access in Ontonym — synced at every sign-in.
Read postWhy personal use is free and unlimited
A memory is only useful when it’s complete. Charging you to fill it would work against the product. You pay when you share.
Read postEvery fact gets a URL
Things in an Ontonym memory aren’t buried in an app. Each one has a readable page you can link in Slack, email, or a doc.
Read post500+ integrations, one directory
Your knowledge lives in the tools you already use. The integrations directory shows how it flows into memory.
Read post