The AI workspace for your audio files

Audio files are the most information-dense, least retrievable format you own. An hour-long interview contains thousands of words of insight, but finding any specific moment means pressing play and scrubbing until you hear it. A lecture recording holds an entire class session, but revisiting one concept means listening to the whole thing again. A collection of ten interviews, five lectures, and a handful of field recordings contains a body of knowledge that's effectively inaccessible the moment the recording stops. You can't search an audio file. You can't skim it. You can't ctrl+F it. The information is there, locked behind a play button.

Fabric transcribes every audio file automatically and makes the transcript searchable by meaning. Find the exact moment across hours of recordings in seconds. Ask the AI questions about what was said. The knowledge in your audio becomes as retrievable as the knowledge in your documents.

Every audio file transcribed automatically

Fabric's audio and video transcription turns every recording into searchable text. Drop in an interview, a lecture, a meeting recording, a field recording, or any audio file, and the transcript is generated automatically. No separate transcription service. No uploading to another tool. No waiting for a human transcriber.

The transcription handles conversational audio, accented speech, and varying recording quality. The transcript is timestamped, so you can jump to the exact moment in the recording from any point in the text.

For recordings you capture yourself, AI voice notes record and transcribe in one step. Record a meeting, an interview, or a thought, and the transcript is in your workspace immediately.

Search inside every recording by meaning

Once audio is transcribed, it becomes as searchable as any document. AI search reads every transcript and searches by meaning. Ask "the part where the participant talked about switching costs" or "when the professor explained the difference between correlation and causation" and find the exact moment from the exact recording across your entire library.

The search doesn't need exact words. It understands concepts. A participant who said "it was too expensive to move to a different provider" is findable by searching "switching costs" even though those words were never spoken. This is the difference between transcript search and meaning-based search, and it's what makes a library of recordings genuinely usable.

Search works across audio transcripts and every other file type in your workspace. A concept from an interview is findable alongside the research paper that addresses it and the notes you wrote about it.

Ask the AI about what was said

The AI assistant treats your audio transcripts as a knowledge base. Ask it to summarise an interview, extract the key points from a lecture, find every mention of a specific topic across multiple recordings, or compare what different people said about the same subject.

For researchers working with interview data, this is transformative. Instead of re-listening to twenty hours of interviews, you ask questions across the full set and get synthesised answers with references to the specific recordings and timestamps. For students revisiting lecture recordings, it means asking "explain the concept from the lecture about X" and getting an answer drawn from the actual lecture content.

The assistant works from your transcripts alongside your other materials, so it can connect what was said in a recording to what was written in a document, a note, or an email.

Annotate recordings with your thinking

Listening to a recording and thinking about it should happen in the same place. Annotations let you mark up transcripts with highlights and comments, the same way you'd annotate a document. Highlight a key quote from an interview. Flag a section of a lecture for revision. Note a follow-up question at a specific point in a meeting recording.

The annotations are searchable. A comment you made on a transcript three months ago is findable by what you wrote. Your thinking about the audio is as retrievable as the audio itself.

Connected to your documents, notes, and research

Audio recordings rarely exist alone. An interview connects to the research project it's part of. A meeting recording connects to the agenda, the follow-up notes, and the decisions made. A lecture connects to the readings and the study notes.

In Fabric, audio files and their transcripts live alongside your documents, PDFs, notes, images, and every other file type. Search returns results from across everything. The interview transcript, the research paper it relates to, and your synthesis notes are findable in the same query. The explorer surfaces connections between recordings and other material, so an interview and a paper that address the same concept are linked even if you didn't connect them manually.

This connectivity is what makes audio useful for research projects, meeting notes, and studying: the recording is part of a larger body of knowledge, not an isolated file.

Capture and organise without effort

Audio arrives from many sources: phone recordings, meeting tools, downloaded files, forwarded attachments. Forward an audio file by email to email-to-note and it's transcribed and searchable. Record directly with AI voice notes on any device. Upload files from your computer. Pull in recordings from Google Drive or Dropbox.

Smart organization automatically tags and categorises audio files by content. Interviews, lectures, meetings, and personal recordings cluster without manual filing. Fabric syncs across devices, so a recording captured on your phone is searchable on your laptop.

Who uses Fabric for audio

Audio is central to many knowledge workflows. User researchers record and search across participant interviews. Researchers and research teams capture field recordings and interviews for qualitative analysis. Students record and search lectures for studying. Writers and journalists record interviews and find specific quotes. Educators record and transcribe lectures and seminars. Lawyers transcribe client calls and depositions. Music creators manage audio references and vocal ideas. Anyone capturing meeting notes from calls and conversations.

For structured approaches to capturing and using audio, see the guide to meeting notes.

Get started

Make every recording searchable by what was said, and stop losing knowledge behind a play button. Try Fabric free.

FAQs

Does Fabric transcribe audio files automatically?

Yes. Audio and video transcription generates a searchable transcript for every audio file. No separate service required.

Can I search for a specific moment across multiple recordings?

Yes. AI search reads every transcript and searches by meaning. Describe what was said and find the moment from the right recording, with a timestamp.

Does the search work on concepts, not just exact words?

Yes. Search by meaning, not by keyword match. A participant who discussed "the cost of changing providers" is findable by searching "switching costs."

Can the AI summarise a recording?

Yes. The AI assistant can summarise any recording, extract key points, or answer specific questions about what was said.

Can the AI synthesise across multiple recordings?

Yes. Ask the assistant a question and it draws from every relevant transcript. Useful for comparing what different interviewees said or finding every mention of a topic across hours of audio.

Can I annotate transcripts?

Yes. Annotations let you highlight and comment on transcripts. Your annotations are searchable by what they say.

Can I record directly in Fabric?

Yes. AI voice notes record and transcribe in one step on any device.

Can I search across audio and other file types together?

Yes. Transcripts are searchable alongside documents, PDFs, notes, images, and everything else. A concept from a recording and a related document are findable in the same search.

What audio formats are supported?

Fabric handles common audio formats including MP3, M4A, WAV, and others. All are transcribed and searchable.

Can I import recordings from Google Drive or Dropbox?

Yes. Fabric connects to Google Drive and Dropbox. Existing recordings are transcribed and searchable once imported.

Are recordings synced across devices?

Yes. Fabric syncs across devices. Record on your phone, search and listen on your laptop.

Are my recordings private?

Yes. Your content is encrypted and only visible to you unless you choose to share it. Fabric uses AES-256 encryption and is CASA Tier 2 compliant.

How is this different from a transcription service like Otter?

Transcription services produce transcripts. Fabric produces transcripts and makes them searchable by meaning across your full library, alongside your documents, notes, and other files, with an AI that synthesises across recordings. The difference is between having a transcript and having a searchable knowledge base built from everything you've ever recorded.