Your Notes Know the Answer — Ask Them

Asking beats searching

Keyword search requires you to remember how you phrased something when you captured it. AI Chat requires you to know what you're trying to learn.

That's a meaningful difference. "What did the client say about the timeline?" is a question you can ask even if you don't remember which meeting it came up in, which document you uploaded, or what you titled the note. Zibri reads across your vault, finds the relevant material, and answers the question directly.

Every answer includes a citation — the specific note it drew from. You can click through, read the original context, and verify that the answer is grounded in something real. If it can't find the answer in your vault, it says so. It doesn't guess.

[PICTURE: screenshot of Zibri AI Chat showing a question, a direct answer, and the source-note citation beneath it]

Scope it to a vault, or open it to everything

AI Chat works at two levels. You can chat with a single vault — useful when you're focused on one client, one project, or one subject area and want answers scoped to that material only. Or you can chat across all vaults at once, which is where cross-cutting questions get answered: "What open commitments do I have across all my client projects?" draws on everything you've put into Zibri.

The vault-scoped version is particularly useful when you've been disciplined about where things live. A client vault that contains meeting transcripts, forwarded emails, uploaded briefs, and your own notes becomes a knowledge base you can interrogate directly. "What requirements did they specify in the initial brief?" is a better question when the vault contains only their material — and Zibri's answer is correspondingly more precise.

Every content type contributes

AI Chat doesn't treat typed notes differently from uploaded documents or forwarded emails or meeting transcripts captured by Zibri's recorder. They all land in the same indexed pool and are all queryable through the same interface.

That means the more you put in — and the wider the variety — the more useful the answers become. A vendor proposal you uploaded, a client brief that arrived by email and got forwarded in, a meeting transcript from the call where you discussed it: all of that becomes context AI Chat can draw on when you ask "What were the terms we agreed to with this vendor?"

This is where Zibri's content pipeline pays off. Email forwarding, the desktop recorder, PDF uploads — each one individually adds notes to your vault. Collectively, they create a knowledge base dense enough that AI Chat can answer questions you'd have no other way of answering quickly.

The questions worth asking

Some of the highest-value uses aren't searches for a specific fact — they're synthesis questions that would take real time to answer manually:

"What has [client name] raised as a concern across our calls this quarter?" Zibri reads across every meeting transcript and note associated with that client and pulls the pattern.

"What open questions came out of last week's planning sessions?" Draws on meeting notes, action items, and follow-up threads — surfaces things you marked as unresolved but haven't acted on.

"What have I read or captured about [topic] in the last month?" Useful before writing a proposal, preparing a presentation, or walking into a meeting where you want to recall what you already know.

"Is there anything I committed to that I haven't followed up on?" Pulls forward the obligations that fell through the gap between notes and task lists.

None of these are retrievable by keyword search in any reasonable amount of time. They're the kind of synthesis question that used to require sitting down and reading through everything yourself.

[PICTURE: screenshot of Zibri AI Chat showing a multi-source synthesis answer, with two or more source citations from different vault notes]

Cited answers are the part that matters

It's easy to build a system that generates plausible-sounding responses. It's harder to build one that's verifiable.

Every Zibri AI Chat answer links to its source. That's not just a feature for skeptics — it's what makes the answer useful in practice. When you ask what was decided in a meeting and Zibri tells you, you can click through to the transcript and confirm it. When you use that answer in a proposal or a client call, you're drawing on something you can actually stand behind.

The citation also does something subtler: it shows you notes you'd forgotten you had. When an answer surfaces a note from six weeks ago that you'd stopped thinking about, the chat has effectively returned something you'd lost — and given you a link to it.

The takeaway

If you've been adding content to Zibri — forwarding emails, recording meetings, uploading documents — AI Chat is where that investment pays out. Open it and ask a question you'd normally have to dig for. The answer will either be there, cited and traceable, or it won't — in which case you'll know exactly what's missing from your vault.

That's more useful than a search bar that makes you do all the work.

— Erin