Explain the problem as you see it
Right now, AI chat in Tana only works with selected nodes. This creates friction:
- It’s often difficult to find the exact node I want to chat with.
- Even if I find it, I could just copy the content into ChatGPT and get the same result.
As a result, the AI chat feature feels disconnected from my knowledge base. It doesn’t provide unique value beyond what I can already do outside Tana.
Why is this a problem for you?
Knowledge are distributed across multiple nodes that are connected in meaningful ways. Insights often come from seeing those relationships, not from looking at a single node in isolation.
I want to use natural language to query my knowledge base, even when I don’t know the exact terms or node to search for.
Without whole-workspace context, AI chat cannot fully support summarization, exploration, or creative insight generation.
Suggest a solution
Introduce a “Chat with Knowledge” mode for AI chat. When enabled:
- The system uses the embedded whole knowledge base as context.
- AI can summarize across multiple related nodes automatically, not just one at a time.
- Users can discover hidden connections, ask broad questions, and receive higher-value insights — making AI chat a truly integrated part of the Tana experience.
Why this matters for Tana:
- It transforms AI chat from a “nice-to-have” tool into a core differentiator of the platform.
- It reduces the gap between knowledge capture and knowledge exploration.
- It makes Tana the place where knowledge not only lives but also actively works for the user.
2 Comments
Hell yeah!
Tana is an absolutely great app, and the data model with supertags is one of the best I’ve ever used.
However, without the ability for AI to search and reason over the entire personal library by default, I don’t see how Tana can stay competitive in the long run. Tools like GetRecall.ai, Granola, or VoiceApp already treat the whole knowledge base as AI context, and this has become a real killer feature.
Global AI access to all notes would take Tana to a completely new level of productivity. With Tana’s speed, structure, and supertags, this could be far more powerful than what competitors offer. I honestly don’t see the advantage of not enabling this.
Granola solves this brilliantly, and a solution similar to Google NotebookLM (which is currently one of the best tools in this space) would be an incredible fit for Tana.
This would be a massive competitive advantage: instant insights, summaries, and answers across your own data, powered by Tana’s unique structure.
Please seriously consider this — it would be a game-changing feature for Tana.