All updates
NewAnalysis & AI

Upload interviews you already recorded and analyze them in Conveo

Bring in interview videos you already have. Conveo transcribes them, separates interviewer from participant and gets them ready for analysis.

Emily Kavanagh

At a glance

  • Upload MP4, MOV, AVI, MKV or WebM videos of up to 6 GB each
  • Choose from more than 40 interview languages before you upload
  • Conveo transcribes each recording and labels interviewer and participant
  • Participant names are picked up from introductions when they are said
  • Run AI analysis, including video insights, on the uploaded interviews

Release details

Shipped
Area
Analysis & AI
Kind
New
Author
Emily Kavanagh
  • Interview analysis
Read the docs guide

Most research teams have interviews nobody has gone back to: the in-depth sessions an agency ran last year, the in-home visits your team filmed, the pilot where half the interviews had a human moderator. Using them again means rewatching hours of video and typing up quotes by hand, so they rarely get compared with anything new.

Interview uploads, how it works and what it changes, no sound needed.

How it works

  • Choose the language. Pick the language spoken in the recordings. Upload recordings in different languages as separate batches.

  • Add your files. Drag in MP4, MOV, AVI, MKV or WebM videos of up to 6 GB each and start the upload. Each interview keeps its original file name as a label.

  • Conveo transcribes and sorts the speakers. Each turn is assigned to the interviewer or the participant, and the participant's name is picked up from the introduction when they say it.

  • Analyze it in Conveo. Once processing finishes, the recordings are ready for the same AI analysis Conveo runs on its own video conversations.

See what happened on camera and on screen

Uploaded studies can also run video insights. Choose the analysis tracks on the study design page, including one that follows participants as they share their screen and move through apps. Every coded moment links back to the clip it came from.

Research teams asked for it in five situations

These come from teams we work with in market research and healthcare research agencies, consumer goods, financial services and technology. We kept them anonymous.

  • Interviews your own team moderated. Sessions run by your researchers or an agency go through the same analysis as your AI-led interviews.

  • Human-led and AI-led side by side. Teams trying AI-led interviews wanted to compare them with human-led sessions on the same discussion guide.

  • The archive nobody reopened. A library of past interviews can be asked a question it was never collected for.

  • Last study next to this one. Recordings from an earlier project sit beside a new study, so earlier findings can be checked against fresh ones.

  • Recordings from the field. In-home visits, ethnography and other filmed sessions get transcribed and coded without a manual pass.

What this means for your research

Recordings you already paid for become usable data again. Transcript, speaker labels and analysis come from one place, so your hours go into checking findings and building the story.

Some limits to know. Uploads take video files, so a transcript document on its own will not process. Speaker labels and names are AI-generated: check them against the recording before you quote anyone. Language is never detected automatically, so select the one actually spoken. Keep the original file until you have reviewed the processed interview. Uploaded studies are in private beta: ask your Conveo team for access.

Read the uploaded studies guide.