--- title: "MeetGeek Integration" description: "Put MeetGeek meeting intelligence into agent workflows. Agents dispatch a recording bot to live calls, submit recordings for AI analysis, and pull transcripts, summaries, highlights, and per-meeting KPIs into downstream tools." url: https://flowrunner.ai/integrations/meetgeekai date_modified: 2026-08-04T10:04:16-07:00 --- # MeetGeek [Productivity](https://flowrunner.ai/integrations/category/project-management-productivity) Put MeetGeek meeting intelligence into agent workflows. Agents dispatch a recording bot to live calls, submit recordings for AI analysis, and pull transcripts, summaries, highlights, and per-meeting KPIs into downstream tools. [Verified](https://flowrunner.ai/integrations/verified "What does verified mean?") · 12 actions · API key · available [MeetGeek website](https://meetgeek.ai/) · [Platform Documentation](https://docs.meetgeek.ai/) · Capability data verified 2026-07-27 1. A discovery call ends and the meeting appears in the account's processed list 2. List Meetings picks up the new meeting ID on the next sweep 3. Get Summary and Get Highlights pull the narrative recap, decisions, and action items 4. The recap posts to the deal channel and the highlights land on the CRM opportunity record 5. Action items labeled next\_steps are matched against existing tasks to avoid duplicates 6. New commitments become tracked tasks assigned to their owners 7. The account executive reviews the AI-extracted commitments before they reach the customer-facing follow-up email, because the summary is a draft of what was promised, not a witness ## What This Integration Enables Every meeting produces two artifacts: what was said, and what everyone now believes was said. MeetGeek captures the first with recordings, transcripts, summaries, highlights, and per-meeting KPIs. FlowRunner closes the gap to the second: agents move the meeting's outputs into the CRM, the task tracker, and the knowledge base while the conversation is still warm, and route the customer-facing pieces through the person who was actually in the room. The result is a meeting record that does work instead of accumulating. - Sweep processed meetings and pull transcripts, summaries, highlights, and insights by meeting ID - Turn extracted action items and decisions into tracked tasks with owners - Archive full transcripts into searchable documents and knowledge bases - Dispatch a recording bot to live Zoom, Google Meet, or Microsoft Teams calls - Submit existing recordings by URL for asynchronous transcription and analysis - Track team meeting activity and effectiveness KPIs across share and view access teams ### Without FlowRunner **The recap depends on the busiest person**: Whoever ran the call owes notes, and the notes arrive Thursday or never **Commitments evaporate**: "We'll send the security doc by Friday" lives only in minute 43 of a recording **Meeting data stays in the meeting tool**: Transcripts, summaries, and KPIs sit in MeetGeek while the CRM stays blank ### With FlowRunner **The recap posts before the next meeting**: Summaries and highlights sweep into the channel and the CRM within the polling interval **Action items become tracked tasks**: Every extracted commitment gets an owner and a due date in the task system **A human signs what leaves the building**: AI drafts the follow-up from the highlights; the person on the call approves it ## Use Case Scenarios ### The sales call that files itself When a demo call finishes processing, the agent pulls Get Summary and Get Highlights, posts the recap to the deal room in [Slack](https://flowrunner.ai/integrations/slack), and writes the decisions to the opportunity record in [HubSpot](https://flowrunner.ai/integrations/hubspot). Highlights labeled next\_steps become issues in [Linear](https://flowrunner.ai/integrations/linear) with owners mapped from participant emails. The AE gets a drafted follow-up email built from the commitments, reviews it against what they remember actually promising, and sends. Total elapsed time from call end to filed, tasked, and drafted: one sweep interval. ### The transcript archive that compounds Every processed meeting flows through a nightly agent: Get Transcript pages through the sentences, and the full text lands in a dated document in [Google Docs](https://flowrunner.ai/integrations/google-docs), indexed by title, participants, and team from Get Meeting. Six months later, when someone asks what the customer said about pricing in March, the answer is a search away instead of a scrub through hours of video. Recording files themselves get fetched through Get Recording Download Link and archived to storage before the temporary links expire. ### Meeting hygiene measured, not lectured An operations agent walks List Teams and List Team Meetings weekly, then pulls Get Insights for each meeting: per-meeting KPIs, discussion themes, and improvement suggestions. The numbers aggregate into a scorecard in [Google Sheets](https://flowrunner.ai/integrations/google-sheets) by team: meeting load, engagement patterns, themes that keep recurring. The ops lead reads trends instead of anecdotes, and the conversation about too many status calls starts from data the calendar cannot argue with. ## Human-in-Loop Highlight Delete Meeting removes the recording, transcript, summary, highlights, and insights in one call, permanently. For a customer conversation, that bundle is often the only record of what was actually said: the pricing caveat, the scope agreement, the objection that explains why the deal stalled. A retention sweep that deletes by age alone will eventually destroy the one transcript a dispute needed. So FlowRunner runs retention as proposal and approval: the agent lists meetings past the retention window, checks each against open opportunities and active projects via Get Meeting's participant and calendar data, and presents the batch with exceptions flagged: "31 meetings past 18 months. 2 are linked to an account in an active renewal. Delete the other 29?" A person decides, and the two exceptions live on. The same instinct governs Send Bot to Meeting: a recording bot joining a live call announces itself to everyone present, so agents dispatch it only to meetings where recording was agreed, never as a silent default. Agent processes routinely Detects exception requiring judgment Clear match Continues automatically Ambiguous Routes to human via preferred channel Human decides Agent resumes with decision ## Agent Capabilities 12 actions ### Meetings 3 - **List Meetings** Retrieves a paginated list of the authenticated user's past meetings, newest first, each with meeting ID and UTC start and end timestamps. The sweep step that feeds every downstream content pull. - **Get Meeting** Retrieves a meeting's full details by ID: title, source platform, host and participant emails, join link, language, timezone, applied template, team IDs, and calendar event ID. The context step behind routing and retention checks. - **Delete Meeting** Permanently deletes a meeting and all associated data: recording, transcript, summary, highlights, and insights. Cannot be undone, and the operation this page's human gate exists for. ### Meeting Content 5 - **Get Transcript** Retrieves the full transcript as paginated sentences, each with speaker name, timestamp, and text. The source for archives and searchable meeting history. - **Get Summary** Retrieves the AI-generated narrative summary and additional insights, available once processing completes. The recap that reaches the channel first. - **Get Highlights** Retrieves AI-generated highlights such as action items, decisions, and key moments, each labeled by category and filterable to a single type like next\_steps. The extraction step behind task creation. - **Get Insights** Retrieves per-meeting performance KPIs, popular discussion themes, and improvement suggestions. The data behind meeting-effectiveness reporting. - **Get Recording Download Link** Generates a temporary, time-limited download link for the original audio or video. Agents fetch the file before expiry when archiving recordings. ### Teams 2 - **List Teams** Retrieves the teams the user can access, split into share\_access and view\_access, each with numeric ID and name. The scoping step for team-level reporting. - **List Team Meetings** Retrieves a paginated list of a specific team's past meetings, newest first, by team ID from List Teams. The sweep behind team-wide archives and scorecards. ### Processing 2 - **Upload Recording** Submits an audio or video recording for transcription and analysis via a publicly accessible URL, with optional language and summary template. Processing is asynchronous; completion is detected by webhook or by polling List Meetings. - **Send Bot to Meeting** Sends a MeetGeek recording bot to join a live meeting by its join link on Zoom, Google Meet, or Microsoft Teams, with optional language, template, and title override. Returns the resulting meeting ID. Dispatched only where recording is consented. ## Frequently Asked Questions ### What can FlowRunner do with MeetGeek? FlowRunner agents can run List Meetings, Get Meeting, and Delete Meeting in MeetGeek, plus 9 more actions. ### Does connecting MeetGeek to FlowRunner require OAuth? No. MeetGeek connects to FlowRunner with an API key, no OAuth flow required. ### Can MeetGeek trigger a FlowRunner workflow automatically? MeetGeek doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger. **Work at MeetGeek?** This integration exposes MeetGeek to AI agents on every FlowRunner plan, including through MCP, at no cost to you. [See what FlowRunner offers integration partners](https://flowrunner.ai/integrations/partners), including how to keep this page current. --- Markdown version of https://flowrunner.ai/integrations/meetgeekai. Site index: https://flowrunner.ai/llms.txt