How to find something someone said in a meeting

You're preparing for a follow-up meeting and you know, with absolute certainty, that someone raised a concern about the timeline three weeks ago. You can almost hear them saying it. But your notes from that meeting are a handful of bullet points about action items, and none of them capture the moment you're thinking of. You scroll through your notebook, your task list, the shared document someone started and nobody finished. Nothing.

This happens constantly, and it happens because meeting notes were never designed to be a complete record. They're a compression of what felt important at the time, written by someone who was simultaneously trying to listen, participate, and decide what mattered. The result is always partial, always biased towards what the note-taker cared about, and always missing the thing you need most three weeks later.

The solution is not better notes. The solution is a searchable record of everything that was said.

Why your meeting notes can't help you

Meeting notes fail for a specific and unavoidable reason: they require someone to make real-time decisions about what matters. The person taking notes has to listen to what's being said, decide whether it's worth recording, condense it into something writable, and do all of this while the conversation continues without them. They will miss things. They will paraphrase in ways that lose nuance. They will skip the tangent that turns out, weeks later, to have been the most important part of the conversation.

Even when notes are thorough, they capture intent rather than language. "Sarah had concerns about the timeline" tells you something happened, but it doesn't tell you what Sarah said, how she framed the concern, or what specific part of the timeline worried her. The difference between a summary and a transcript is the difference between knowing something was discussed and knowing exactly what was said.

This is why meeting notes work best as a companion to a full record, not a replacement for one. Your notes capture your interpretation and your priorities. The transcript captures everything else.

Recording without the awkwardness

One reason people don't record meetings is the tooling. Traditional meeting bots join your call as a visible participant, announcing their presence to everyone and changing the dynamic of the conversation. People speak differently when they know a bot is recording them. They hedge more, say less, and save their honest observations for after the call ends.

Fabric records meetings directly, with automatic transcription and no meeting bot required. The recording happens on your end, which means there's no robotic participant joining the call, no notification that makes everyone self-conscious, and no third-party service storing your conversation. For in-person meetings, you can record and transcribe just as easily using a voice note.

The transcription runs automatically once the recording finishes. Every word, attributed to the person who said it, marked with a timestamp. The full text becomes searchable immediately, which means every meeting you record becomes a permanent, searchable part of your knowledge base. Because Fabric acts as a persistent layer that indexes everything across your connected tools, meeting transcripts sit alongside your documents, emails, and notes in one searchable space.

Searching with timestamp precision

When you need to find something someone said, the most direct approach is simply to search for it. Fabric searches inside audio and video transcriptions, so you can type a phrase, a name, a topic, or even a rough approximation of what was said, and find matching moments across every meeting you've recorded.

The results come back with clickable timestamps. Each result shows you the surrounding context, so you can see what was said before and after the moment you're looking for, and clicking the timestamp jumps you directly to that point in the recording. You hear the actual words in the actual voice of the person who said them. There's no ambiguity about what was said or how it was said.

This works even when you can't remember the exact phrasing. If you search for "timeline concern" and someone said "I'm worried we won't hit the February deadline," the search is smart enough to surface relevant moments rather than demanding an exact string match. You can also search across all your apps at once, which means if the meeting spawned follow-up emails or documents, those will appear alongside the transcript results.

Asking the AI what was said

Sometimes you don't know what to search for. You remember the shape of the conversation but not the words. You know it was about budgets, or headcount, or that thing the client mentioned about their European expansion, but you can't construct a search query precise enough to find the right moment.

This is where the AI assistant becomes essential. Because Fabric's assistant has access to your transcribed content with timestamp precision, you can ask it questions in natural language. "What did Sarah say about the timeline in last month's project reviews?" or "Which meeting discussed the European expansion?" or "Find the conversation where someone pushed back on the Q2 targets."

The assistant searches across your meetings, finds the relevant moments, and gives you answers with references back to the specific timestamps. You can verify everything it tells you by clicking through to the original recording. This is particularly powerful when you're trying to find recurring themes across your meetings, because the assistant can synthesise information from multiple conversations rather than making you search them one at a time.

You can even ask AI questions during a live meeting, pulling in context from previous discussions while the current one is still happening.

Combining transcripts with your own notes

Transcription doesn't make note-taking pointless. It makes it better. When you know the full conversation is being captured, your notes can focus on what a transcript can't provide: your reactions, your interpretations, the connections you're drawing to other projects, the questions you want to raise later. You stop trying to be a court reporter and start thinking.

Fabric lets you combine your meeting notes with the transcript, merging what you wrote during the meeting with the full record of what was said. Your annotations sit alongside the transcribed text, so when you return to the meeting later, you see both the objective record and your subjective response to it. This combination is far more useful than either one alone.

The same principle applies to voice notes. Quick spoken thoughts recorded before or after a meeting get transcribed and become searchable alongside everything else. That insight you had walking back from the conference room, the one you captured in a thirty-second voice memo, becomes findable months later when you need it.

Building a searchable meeting archive

The real power of this approach reveals itself over time. Each meeting you record adds to a growing archive of searchable conversation. After a few months, you have something remarkable: a complete, searchable history of every important discussion your team has had.

This changes how you prepare for meetings. Instead of relying on someone's summary or your own fading memory, you can search the actual record. You can trace how a decision evolved across multiple conversations. You can find the moment someone first raised an idea that later became a major initiative. You can settle disagreements about what was agreed by going back to the source.

It also changes how you take notes in the first place. When you trust that the system will capture and preserve everything, you write differently. Your notes become more analytical, more forward-looking, more useful. You spend less time documenting what happened and more time thinking about what it means. Fabric's memory engine deepens this over time, building context about recurring topics and participants so that the archive grows more useful the longer you use it.

The meetings you've already recorded are already searchable. The question someone asked three weeks ago, the one you can almost hear but can't quite place, is sitting in your archive right now. You just need to look for it.

Frequently asked questions

Can I search for something if I don't remember the exact words?

Yes. Fabric's search handles approximate queries and semantic matching, so you can search by topic, concept, or rough phrasing rather than needing the precise words someone used. The AI assistant is even more flexible, accepting natural-language questions like "when did we discuss the budget shortfall" and returning relevant moments from across your meetings.

How quickly after a meeting can I search the transcript?

Transcription runs automatically once the recording finishes, and the process typically completes within minutes. The exact time depends on the length of the meeting, but even hour-long recordings are usually searchable shortly after they end.

Does the transcription identify who said what?

Yes. The transcription includes speaker attribution, so you can see which participant said each part of the conversation. This makes it possible to search for what a specific person said, or to ask the AI assistant questions like "what did the client say about pricing."

What if the meeting was in person, not a video call?

Fabric handles in-person meetings through its voice note and recording features. You can record an in-person meeting directly from Fabric, and the transcription and search work exactly the same way. See the guide on recording and transcribing in-person meetings for the details.

Can I search across meetings and other content at the same time?

Yes. Fabric's search works across all your content, including meeting transcripts, notes, documents, and files from connected apps. This means a single search can surface the meeting moment where something was discussed alongside the follow-up email or document that resulted from it.

How do the clickable timestamps work?

Each search result from a transcription includes a timestamp that corresponds to the moment in the recording. Clicking the timestamp jumps you to that exact point in the audio or video, so you can hear (or watch) what was said in its original context. The surrounding transcript text is visible too, so you can read ahead or behind the matched moment.

Will this replace the need for meeting summaries?

Not entirely. Summaries and auto-generated meeting summaries are still useful for getting a quick overview of what a meeting covered. The difference is that when a summary leaves out something important, or when you need the exact words rather than a paraphrase, you can go straight to the transcript. The two work together rather than one replacing the other.

Can I use this for meetings recorded on Zoom or other platforms?

Fabric can transcribe Zoom meetings without a bot, recording on your end without any third-party participant joining the call. The same approach works across other video conferencing platforms. Everything gets transcribed and becomes part of your searchable archive.

What if the audio quality is poor?

Transcription accuracy depends on audio quality, but modern transcription handles most real-world conditions well, including background noise, crosstalk, and varying microphone quality. For critical meetings where you want the best possible transcript, using a decent microphone and minimising background noise will help. Even imperfect transcriptions are searchable and usually capture enough to find the moment you're looking for.

How far back can I search?

There is no time limit on searching your meeting archive. Every meeting you've recorded and transcribed remains fully searchable indefinitely. The archive grows with each meeting, and older recordings are just as searchable as recent ones.

Related pages

• How to search across all your apps at once

• How to find something you saved months ago

• How to auto-generate meeting summaries

• How to transcribe Zoom meetings without a bot

• How to record and transcribe an in-person meeting

• How to combine your meeting notes with the transcript

• How to ask AI questions during a live meeting

• How to find recurring themes across your meetings