Summarise customer interviews with AI
Turn an authorised transcript into a clear research note covering current workarounds, satisfaction, unexpected findings and agreed actions.
Summarise customer interviews with AI while the details are still fresh. A shop owner who records stock on paper at closing time may value being able to work without a connection, even while disliking the next day's data entry. The summarize-interview skill helps preserve both sides of that account in a structured note.
Share a transcript you are allowed to process. The agent can separate the customer's current approach, what works well, problems, unexpected observations and agreed actions. When the person gave no date or score, leave that fact open. A guess that looks precise can easily become a false requirement later.
What it does
- Describe the customer's context and the job they need to finish.
- Record what the existing solution already does well.
- Connect a problem to an example in the participant's words.
- Capture an unexpected finding without claiming it represents every customer.
- List follow-up work with a date and owner where the conversation establishes them.
Install and use
- Install the skill in your agent and provide a permitted transcript. Remove unnecessary identifiers before sharing. No separate transcription service is included.
- Ask for a research note that distinguishes the participant's statement from the team's interpretation.
- Check the summary against the original, then store the Markdown document with the appropriate access.
FAQ
Source
Made by Paweł Huryn. Original skill: summarize-interview, published on GitHub under the MIT licence.
View on GitHub