Overview
The tl;dv MCP allows AI assistants to securely access and analyze your meeting recordings and transcripts directly from the tl;dv workspace.
It will allow your favourite LLM to actually work with all the intelligence and data you have gathered from your meetings and that live inside your tl;dv workspace, instead of working in separate silos.
Turn your conversations into structured insights that you can query in natural language, helping you move faster across sales, customer success, and internal collaboration.
Setup Instructions
Requirements
Claude - Paid plan (Pro, Max, Team, or Enterprise) + access to Claude Connectors
ChatGPT - Plus, Team, or Enterprise account + access to ChatGPT Connectors
tl;dv - Pro Plan (or higher)
Steps
In your tl;dv workspace, go to Integrations → Marketplace. From there, you can manage integrations and access the connector setup process.
Search for tl;dv (or tldv without the semicolon)
Click Install, then Connect
Authenticate your tl;dv account
Grant access permissions
Once connected, you can start querying your meetings immediately, or connect more tools for better intel and actions.
⚠️ ChatGPT users: your workspace admin may need to enable Connectors before this step is available.
What You Can Do
1. Search & Summarize Meetings
Instantly extract insights from your past meetings without rewatching recordings, and take these learnings to where you’re actually working.
Examples:
“Summarize all my action items from the past 7 days and create a database in Notion”
“What decisions were made in product meetings this week? Compare them to the engineering stand-up meetings and give me a timeline for planned releases ”
“Give me a summary of all meetings with ACME Corp and compare it to the Contact notes in Hubspot”
“What did we promise the client in our last call? Give me a list of pain points the client indicated, analyze our roadmap from Jira and give me a comparison table between the two”
Outputs:
Action items in a digestible format
Decisions & key takeaways in context
Meeting summaries across time ranges and teams
Intel from calls analyzed with the context of data from other tools
2. Score Sales Calls Automatically
Evaluate your team’s performance against structured sales frameworks.
Examples:
“Score all sales calls this month against MEDDICC”
“Which calls had a clearly identified economic buyer?”
“Show me deals where no pain was identified”
“Compare top-performing reps vs low-performing reps based on call quality”
Outputs:
Call scoring per framework (e.g. MEDDICC)
Strengths & gaps across reps
Coaching insights and areas for improvement
3. Generate High-Quality Follow-Ups
Turn any meeting into a polished follow-up email in seconds.
Examples:
“Summarize my last meeting with ACME Corp and send a follow-up email”
“Draft a recap email with next steps and deadlines and send it to my team lead”
“Write a friendly follow-up to the prospect highlighting open questions”
Outputs:
Structured recap emails
Professional tone aligned with context
Automation of post-meeting admin work
4. Voice of the Customer (VoC) Insights
Analyze all customer conversations to uncover patterns, risks, and opportunities.
Examples:
“Analyze all of the external calls of our CS reps, create a Voice of the Customer report from the last 30 days and publish it in Confluence as a new page”
“What are the top complaints mentioned by customers and cross-check our Jira boards for similar ongoing or known issues?”
“Which accounts show churn risk based on recent calls and confirm what MRR they represent from our RevOps files?”
“Identify upsell opportunities across customer conversations and send a Slack message to the CS reps with the highlights”
“What features are customers requesting most often? Compare to our publicly available feature request board and where they coincide provide a comparison table.”
Outputs:
Aggregated customer insights
Churn signals and pro-active retention measures
Upsell opportunities with quick signaling
Feature request trends across different platforms
5. Get Creative - Discover the Magic ✨
The real power of the tl;dv MCP comes from combining prompts & tools in creative ways.
Don’t hesitate to experiment; some of the most valuable use cases come from unexpected questions.
Try things like:
“Act as a VP Sales and audit our pipeline health based on recent calls”
“What patterns do you see across our lost deals?”
“Summarize how our messaging has evolved over the past 3 months”
“Identify contradictions between what we promise and what customers say”
“If you were a competitor, what weaknesses would you spot in our calls?”
“Turn all customer objections into a battlecard for the sales team”
We’re especially excited to see how you use this! Feel free to push the limits and explore what’s possible.
How It Works
You connect your tl;dv account via the connector marketplace of Anthropic or OpenAI
The AI assistant securely accesses your meeting transcripts and notes
You ask questions or provide instructions in natural language
Insights are generated instantly from your conversations
Your instructions are performed by the assistant across the tools you have connected
No manual tagging, filtering, or note-taking required.
Best Practices
To get the most accurate results:
Use timeframes (“past week”, “last 30 days”) and as a rule of thumb, the number of meetings returned will always be 50.
Be explicit about your goal (“summarize”, “analyze”, “score”, “generate”)
Reference titles, emails or teams as much as possible.
The connection with tl;dv focuses on meetings you’ve participated in by default. To expand your results to all meetings you can access, just adjust your prompt, for example: “include meetings I wasn’t part of”.
Troubleshooting
No Results Returned
Ensure you have recorded meetings with available transcripts
Check that the correct tl;dv account is connected
No access to a specific meeting
Ensure the organizer of that meeting is also a Pro user (or higher)
Missing or Incomplete Insights
Try narrowing your query (e.g. specific timeframe or company)
Ensure meetings fall within your requested period
Connection Issues
Reconnect your tl;dv account from the connector settings
Refresh your session or restart your AI client
Support
If you need help:
Email: support@tldv.io
Chat: Use the Intercom widget inside the tl;dv app
Summary
The tl;dv MCP will allow your AI assistant to connect all the valuable intel that already lives in your calls and your tl;dv ecosystem to so many of your favorite tools, giving it more context and enriching it further. It allows you to:
Save hours reviewing recordings and doing admin work
Improve sales performance by giving you direction on how to coach your reps
Automate follow-ups and save time and team capacity on them
Understand your customers’ needs and feedback at scale
Remove the silos that your company is operating in
All through simple natural language queries.