# Meeting or Call Transcript Summarizer

Summarizes a raw meeting or call transcript into key discussion points, decisions, and open questions, filtering out filler and cross-talk without losing important nuance.

## Prompt

You are a research assistant who specializes in turning messy raw transcripts into clean, useful summaries.

Raw transcript: {{raw_transcript}}
Meeting/call purpose: {{meeting_purpose}}
Who was involved (names/roles if known): {{participants}}

Do the following:
1. Write a 3-4 sentence high-level summary of what the meeting/call was about and its general outcome.
2. List the main discussion topics in the order they came up, with a 1-2 sentence summary of what was actually said about each, filtering out filler words, cross-talk, and repeated restatements.
3. Separately list any decisions that were explicitly made during the conversation.
4. Separately list any open questions or unresolved points that came up but weren't settled.
5. Note any strong disagreements or notably different opinions expressed, since these are often important context that gets lost in a purely neutral summary.

## Best for

Professionals reviewing recorded calls, interviews, or meetings who need a clean, structured summary from a raw or auto-generated transcript full of filler, interruptions, and unclear attribution.

## Compatible tools

- Claude
- ChatGPT
- Gemini

## How to use

- Paste the raw transcript, even if it's messy with filler words, cross-talk, or unclear speaker labels.
- Describe the meeting's purpose and who was involved for more accurate context in the summary.
- Use the decisions and open questions sections as your action-tracking reference.
- Pay attention to any flagged disagreements or tone shifts, since these are often the most important but easiest to miss details in a raw transcript.

## Customization tips

- If speaker labels in the raw transcript are unclear or generic (Speaker 1, Speaker 2), provide context on who's likely who based on what they said, to improve attribution accuracy.
- For sales or client calls specifically, ask for an additional section flagging any buying signals or objections mentioned.
- For very long transcripts, paste in sections and ask for the summary to build cumulatively across each part.

## Example input

Raw transcript: [a messy 40-minute auto-generated transcript of a client discovery call with cross-talk and filler words]. Meeting purpose: Understand a potential client's needs before proposal. Participants: Sales rep (me) and prospective client, a marketing director.

## Example output

A high-level summary noting the client is exploring options for automating their content approval workflow, main discussion topics covering current pain points, budget range, and timeline expectations, a decisions section noting the client agreed to review a proposal by next Friday, an open questions section flagging that the exact number of team members who'd use the tool was never clarified, and a note on a disagreement where the client's tone shifted noticeably skeptical when pricing was first mentioned, worth following up on directly.