Turn meeting notes into reliable CRM records by giving Claude a tightly scoped Zapier MCP toolset and project instructions that define lookup order, field meanings, and when to create or update a record.
Source episode · 05:00
How I AI: Reid Robinson's Zapier Workflows for CRM Automation, Meeting Prep, and Feedback Loopswith Claire Vo
Reid creates a restricted Zapier MCP tool collection, teaches a Claude Project the exact CRM lookup and write sequence, and processes meeting notes into Coda or HubSpot.
An updated CRM record with the correct contact, meeting, next steps, opportunity details, and source transcript.
Workflow automation platform
Anthropic AI assistant
AI-powered meeting note-taker and summarizer
Step by step
Follow the sequence once, then adapt the prompts, checks, and handoffs to your own setup.
3 steps
Create a dedicated Zapier MCP server for CRM logging. Add only the required Coda, HubSpot, calendar, internal lookup, and note actions, and restrict each tool to the records or documents this flow may touch.
Create a Claude Project that specifies the lookup sequence, record identity rules, field definitions, allowed values, and the exact conditions for create versus update.
For every meeting transcript: 1. identify the attendee and meeting date; 2. search the authorized CRM by email; 3. search the meeting table for the same contact and date; 4. use the internal lookup only when required; 5. update the matching record or create one only when no match exists; 6. populate Next Steps, Opportunity Details, Product Feedback, Owner, and Source using the attached field definitions; 7. report every write and any unresolved ambiguity.
After the meeting, select the CRM Logging project and attach the transcript plus any meeting identifier. Let Claude run the authorized lookup and write tools, then review its change summary against the source notes.
Process this meeting transcript using the CRM Logging project. Search before creating records, preserve the source link, update the correct contact and meeting, and report the exact fields changed. Ask before writing if identity or field mapping is ambiguous. [transcript]
Turn an idea into a PRD, user stories, and a plan.
After the steps
How to recover when the loop fails and where human judgment helps.
Recover
Keep building
Automate and accelerate executive-level tasks by using AI agents integrated with your business tools. Learn how to use prompts to redesign Figma slides, analyze sales calls for insights, and generate lead lists from your email.
Go beyond simple automation by creating a system that measures its own performance and improves itself. Learn to track metrics, score AI agent runs to find failures, and use observer agents to automatically fix your factory's code.
Learn how to automate your software development lifecycle, turning a simple Slack request into a fully implemented, tested, and QA-verified pull request in GitHub, reflecting a true software factory model.
Join 100,000+ product managers who use ChatPRD to write better docs, align teams faster, and build products users love.