Otter.ai MCP

@otter.ai pull the transcript from yesterday's standup meeting and highlight any action items mentioned

Retrieved transcript from 'Daily Standup - Jan 15' (14:32 duration). The meeting included 4 participants. Action items identified: Sarah committed to finishing the API documentation by Friday; Mike will review the staging deployment today; Team agreed to schedule a retro for next Tuesday. Full transcript shows discussion of sprint progress, a blocker with the payment gateway, and clarification on the new feature spec.

This example shows transcript retrieval and AI-assisted parsing. Otter.ai requires OAuth consent with read access to your meeting recordings. The quality of action-item extraction depends on how clearly commitments were stated during the meeting — vague discussions may not surface as discrete tasks.

@otter.ai search all my transcripts from the last two weeks for mentions of 'budget approval' and summarize what was decided

Found 3 meetings mentioning 'budget approval': (1) Jan 8 Leadership Sync — CFO approved Q1 marketing spend up to $45K, pending final deck review. (2) Jan 10 Product Planning — Budget for contractor hours deferred until Feb board meeting. (3) Jan 14 All-Hands — No new budget discussion, brief mention in Q&A about hiring freeze. Consensus: marketing approved, contractor budget still pending.

Demonstrates cross-meeting search and synthesis. This requires OAuth access to your full Otter.ai library. Search accuracy depends on transcription quality — accents, background noise, or jargon may reduce match precision. Results reflect meetings you have access to, not org-wide recordings unless shared.

@otter.ai create a summary of today's client call and share it with the #sales channel format

Generated summary for 'Acme Corp Discovery Call - Jan 16' (42 min). Key points: Client needs integration with Salesforce and Slack by March; current pain point is manual data entry across 3 tools; budget range $20-30K annually; decision-maker is VP Ops, final approval requires CFO sign-off. Next step: send proposal by Jan 23. Summary formatted for Slack sharing with participant names and timestamps for key quotes.

Shows post-meeting summarization and formatting. This example assumes the MCP can generate summaries from recent transcripts — actual implementation may require you to specify the meeting by ID or timestamp. Sharing functionality depends on whether the MCP supports direct posting or just returns formatted text for manual distribution.

When Otter.ai MCP works for async meeting follow-up

A 6-person remote team runs three Zoom calls a week and needs action items surfaced without manual note-taking. The Otter.ai MCP connects via OAuth2, so any team member can query transcripts directly in Switchy without leaving the workspace. The win is speed: instead of scrolling through Otter's web UI or asking "who said what about the API deadline," you ask Switchy and get timestamped excerpts in seconds. The trade-off is scope—if your team records calls outside Otter (Slack huddles, in-person whiteboard sessions), this MCP won't see them. It's the right call when Otter is already your single source of truth for meeting audio and you want that context available in every AI conversation your team has.

Using Otter transcripts to build feature briefs

A 4-person product team conducts 8-10 customer interviews per sprint and needs to turn raw transcripts into feature briefs without re-listening to hours of audio. The Otter.ai MCP lets Switchy pull transcript data directly, so you can ask "summarize pain points from last week's interviews" and get a synthesized view across multiple calls. The advantage is consolidation—no copy-pasting from Otter into Notion or Google Docs. The limitation is that the MCP doesn't yet expose speaker labels or sentiment tagging (as of this input snapshot), so nuanced analysis still requires manual review. This setup wins when your interview volume is high enough that manual synthesis is a bottleneck but low enough that you're not building a dedicated research ops pipeline.

When Otter MCP helps new reps learn from recorded calls

A 3-person sales team onboards a new rep and wants them to learn objection-handling patterns from the last 20 discovery calls without sitting through full recordings. The Otter.ai MCP gives Switchy access to those transcripts, so the new rep can ask "show me how we handled pricing objections in Q1" and get relevant excerpts with context. The benefit is searchability at scale—Otter's native search is good, but querying across calls in natural language inside Switchy is faster for pattern-spotting. The boundary is coaching depth: if you need call scoring, talk-time ratios, or CRM integration, you're better off with a dedicated sales intelligence tool. Use this MCP when your coaching workflow is transcript-first and your team already lives in Otter for call recording.