How to record and transcribe an in-person meeting

In-person meetings have a problem that video calls solved years ago. When you sit across a table from someone, there is no transcript button. No bot joins the room. Nobody is quietly converting speech to text in the background. You leave the conversation with a head full of ideas and a notebook full of fragments, and within a few hours the details start to fade.

The usual workarounds are not great. Typing notes while someone is talking splits your attention. Assigning a dedicated note-taker means one person misses the discussion. And asking everyone to join a video call from the same room, just so a transcription bot can listen in, defeats the purpose of meeting face to face.

What most people overlook is that you already carry a high-quality microphone everywhere you go. Your phone can record an entire meeting, workshop, or one-to-one conversation with a single tap. The missing piece has always been what happens after the recording: getting a usable transcript, pulling out the important points, and connecting those notes to everything else you are working on. That is the part Fabric's voice memo feature handles.

How it works in practice

Before the meeting starts, open Fabric on your phone and tap to begin a voice memo. Place your phone in the centre of the table, or keep it in front of you if it is a one-to-one conversation. Then forget about it. Give your full attention to the people in the room.

When the meeting ends, stop the recording. Fabric uploads the audio to your cloud drive and begins transcribing it automatically. You do not need to export the file, open a separate transcription tool, or wait at your desk. The transcript is ready by the time you get back to your computer, sometimes sooner.

This is worth contrasting with how online meeting transcription typically works. If you have used bots that join video calls, you know the routine: a notification pops up, participants feel watched, and the bot sometimes drops out mid-call. Recording an in-person meeting with your phone avoids all of that. There is no bot to configure, no calendar integration to set up, and no awkward moment when an automated attendee announces itself. For a deeper look at bot-free meeting capture, see the guide on how to take meeting notes without a bot joining the call.

Beyond the transcript

A raw transcript of a forty-minute meeting is useful but not exactly convenient. It might run to several thousand words, and reading through it takes almost as long as the meeting itself. This is where Fabric's AI assistant changes the shape of the work.

Once the transcript is ready, you can ask the assistant to summarise the meeting, extract decisions, list open questions, or pull out anything else that matters to you. The queries are natural language, so you are not limited to predefined templates. You might ask "what did Sarah suggest about the timeline?" or "list every action item with the person responsible" and get a clear, specific answer drawn from the full transcript.

This turns a recording from a static archive into something you can interrogate. Weeks later, when you cannot remember whether a particular detail was agreed or just discussed, you can search for it rather than scrubbing through audio. Fabric's semantic search works across all your recordings, so you do not need to remember which meeting the detail came from. You search by meaning, not by filename.

Capturing more than audio

In-person meetings often produce artefacts that a voice recorder cannot capture. Someone draws a diagram on a whiteboard. A printed agenda gets passed around with scribbled annotations. Handwritten notes pile up in front of every participant.

Fabric's quick capture lets you photograph whiteboards, handouts, or anything else on the table during the meeting. Those images are stored alongside your voice memo, so the full context of the discussion stays together. If you tend to take handwritten notes during meetings, the handwriting scanner can digitise your pages and make them searchable. This is particularly useful for workshops where the thinking happens on paper and whiteboards as much as in conversation.

The result is a complete record of the meeting: audio, transcript, photographs, and handwritten notes, all in one place, all searchable, all connected.

Turning talk into tasks

The point of most meetings is to decide what happens next. The problem is that action items agreed in conversation have a tendency to evaporate. Someone was going to send a proposal. Someone else was going to check availability. A week later, nobody is sure who committed to what.

After Fabric transcribes your recording, you can ask the AI assistant to extract action items automatically. Each item can become a task in Fabric with a reminder attached using tasks and reminders, so follow-ups do not depend on memory alone. The chain from spoken word to tracked task happens without switching between apps or copying text from one tool to another.

How the memory engine ties it together

One of the less obvious benefits of recording meetings in Fabric is what happens over time. Each transcript, each set of notes, each captured whiteboard photograph feeds into the memory engine. Fabric builds connections between your recordings and everything else in your workspace: documents, bookmarks, notes, previous meeting transcripts.

This means that when you are preparing for a follow-up meeting, Fabric can surface what was discussed last time, what tasks were created, and what related material you have saved since then. The context builds up gradually and automatically. You are not filing things into folders or tagging them manually. The connections emerge from the content itself.

For people who have a lot of in-person conversations, whether client meetings, research interviews, project check-ins, or workshops, this accumulation of searchable, connected context becomes remarkably valuable. Six months of recorded meetings form a knowledge base that no set of handwritten notes could match.

A note on recording etiquette and consent

Before you record any conversation, tell the people in the room. In many jurisdictions, recording without consent is illegal. Even where it is legal, it is poor practice. A simple "I would like to record this so I can focus on the conversation rather than taking notes, is that all right with everyone?" is usually enough. Most people appreciate not having to take notes themselves.

If anyone objects, respect that. You can still use Fabric to capture your own spoken reflections after the meeting, dictating your thoughts and key takeaways while they are fresh. The guide on how to keep notes by talking instead of typing covers this approach in more detail.

Privacy and security

Meeting recordings often contain sensitive information: commercial discussions, personnel matters, strategic plans. Fabric encrypts your data with AES-256 encryption and holds CASA Tier 2 certification. Your recordings are stored in your private cloud drive, not shared with third parties for model training or any other purpose. The privacy and security architecture is designed so that your voice memos receive the same protection as any other confidential document.

Working with your recordings through other tools

If you already use ChatGPT, Claude, or other AI tools in your workflow, Fabric does not ask you to abandon them. Through MCP and agent access, your AI tools can reach into Fabric to retrieve meeting transcripts, search across recordings, or pull context from your memory engine. Access is scoped and time-limited, so you control what each tool can see and for how long. Your recordings become part of a broader knowledge layer rather than sitting in an isolated silo.

Frequently asked questions

How good is the transcription quality for in-person recordings?

Transcription quality depends on audio clarity. A phone placed on a table in a quiet room with two or three people produces very accurate transcripts. Larger groups, background noise, or significant distance from the phone will reduce accuracy. For best results, place the phone centrally and ask participants to speak at a normal volume. Fabric's transcription handles multiple speakers and varied accents well, though extremely noisy environments like busy cafes will produce more errors.

Can I record meetings that last several hours?

Yes. Fabric's voice memo feature supports long recordings, and the file is uploaded to your cloud drive, so you are not limited by your phone's local storage in the long term. For very long sessions like full-day workshops, you may want to pause and restart between segments. This makes it easier to find specific discussions later.

Does it work offline or do I need an internet connection to record?

You can record without an internet connection. The audio is captured locally on your phone. Transcription happens once the recording is uploaded, which requires a connection. So you can record in a basement meeting room with no signal and the transcription will begin as soon as your phone reconnects.

How quickly is the transcript available after I stop recording?

Transcription typically completes within a few minutes for a standard meeting-length recording. Longer recordings take proportionally longer. In most cases, the transcript is ready by the time you walk back to your desk.

Can I share a transcript with someone who was not at the meeting?

Yes. You can share transcripts, summaries, or extracted action items from Fabric. This is useful for keeping absent colleagues informed without asking them to listen to an entire recording.

What if the meeting switches between languages?

Fabric's transcription supports multiple languages. If your meeting moves between languages, the transcription will follow, though very rapid switching mid-sentence may reduce accuracy. For meetings conducted primarily in one language with occasional phrases in another, the results are generally reliable.

Is there a way to highlight or annotate specific parts of the transcript?

You can ask the AI assistant to identify and extract specific sections of the transcript. You can also create notes or tasks linked to particular points in the discussion. The semantic search feature means you can always find a specific moment by searching for what was said rather than scrolling through the full text.

How does this compare to using a dedicated voice recorder?

A dedicated recorder captures better audio in some conditions, but it creates a file you then need to transfer, upload, and process separately. Using your phone with Fabric keeps everything in one workflow: recording, transcription, summarisation, task creation, and long-term memory all happen in the same system. The convenience of having everything connected usually outweighs any marginal audio quality difference.

Can I use this for interviews or research conversations?

Recording and transcribing interviews is one of the strongest use cases. Researchers, journalists, and hiring managers all benefit from being able to focus on the conversation and then search across multiple interviews by topic afterwards. The memory engine is particularly useful here, as it surfaces connections between interviews that you might not notice on your own.

What file formats are the recordings saved in?

Recordings are stored in your Fabric cloud drive in standard audio formats. You do not need to manage file formats or conversions. The transcript is available as text within Fabric and can be used with the AI assistant, searched, or shared directly.

Related pages

• AI voice notes

• Audio and video transcription

• How to take meeting notes without a bot joining the call

• How to keep notes by talking instead of typing

• How to turn meeting notes into action items automatically