.avif)
Why this GTM Ops Director uses Grain
TL;DR
- Jess Herbert, Director of GTM Operations at Canvas Medical, an EMR platform, built a daily executive readout that pulls transcripts from every company meeting into Notion and surfaces wins, risks, and sentiment patterns for leadership.
- She also built a sales-to-implementation handoff that turns raw discovery call transcripts into implementation-ready briefs, automatically, using the Grain MCP.
Why most AI meeting tools sit unused
Most teams treat their meeting recordings like email archives: skimmed once when something comes up, then left to pile up. AI summaries get read on the day of the call and never again.
The teams getting real value from meeting intelligence aren't searching the archive. They're running it as a live data source. Every transcript feeds something: a leadership brief, an implementation handoff, a customer story library.
That's what the Grain MCP makes possible. When Grain connects to Claude, your entire call history becomes queryable on demand, and workflows can trigger on a schedule. That's what Jess Herbert built.
What are Grain MCP workflows?
Grain MCP workflows are automated processes that connect the Grain Model Context Protocol to Claude, letting you pull meeting transcripts or clips, and route structured outputs to Slack, Notion, Linear, and email, in plain language, without writing code. You describe what you want; the MCP handles retrieval and Claude handles processing.
The value that's hard to quantify
Jess Herbert runs GTM operations at Canvas Medical, a programmable EMR platform. Her scope covers marketing, GTM systems and automation, and customer intelligence. Her company runs on complex customer calls: healthcare organizations with specific clinical workflow requirements, implementation cycles that span weeks to months, and senior leadership that needs visibility without sitting on every call.
When she started using the Grain MCP, the first measure was time. A few minutes per transcript, a few summaries written faster. Decent, but not the point.
"It's not only that it saves me time," Jess said. "It's how the MCP compiles everything together to create that context that is just priceless. A human couldn't aggregate the context at the same level."
The compiled context across dozens of calls (who said what, what changed, what keeps coming up) is something no person can hold in their head.
Workflow 1: Jess's Daily Executive Readout
What it is: A daily automated summary that reads across the company's meeting transcripts and surfaces wins, risks, and sentiment patterns for senior leadership.
The flow:
- Grain captures transcripts from all company meetings
- Transcripts are stored in a database as meetings happen
- The daily job queries that stored history and generates a structured readout
- The readout surfaces: wins from the period, emerging risks, and sentiment patterns across customer calls
- Leadership reads the brief without attending calls or watching recordings
What it produces:
A daily digest of what happened across every company meeting. It includes risks that need attention, positive signals worth celebrating, and trends in how customers are feeling.
"We do a daily meeting summary," Jess said. "It's an executive readout across all of our meetings. We take all of the transcripts from calls that day across all teams and store them in a database like a Notion database, and then we have our internal Claude agent go and summarize the content — and it brings in any wins, good sentiment, any risks, and also sentiment patterns."
Copy-paste this prompt to build it:
Every day, pull the transcripts from all of our company meetings in Grain.
Then summarize into one executive readout for senior leadership. Include:
- Wins: positive outcomes, resolved issues, or good news from any call
- Risks: anything that sounds like a blocker, complaint, or churn signal
- Sentiment patterns: how customer tone is trending across recent calls
Keep it short enough that a leader can read it in under two minutes. Write it to [location] every morning.
Why this is worth copying:
Senior leaders at growing companies are often the least informed about what customers say simply because they're least likely to be on the calls. This workflow closes that gap without adding meetings to anyone's schedule. If you have a Notion workspace and Grain + Claude, the infrastructure is already there.
Workflow 2: Jess's Sales-to-Implementation Handoff
What it is: An automated briefing that takes raw discovery call transcripts from sales and restructures them into the format the implementation team needs, cutting the manual handoff work substantially.
Additional Context: Canvas Medical is a care modeling platform. Sales discovery surfaces highly specific clinical workflow requirements. Implementation needs that information organized around what they care about, not sorted chronologically by when it was said.
The flow:
- Sales reps conduct discovery calls; Grain captures the transcript
- Jess's skill pulls the raw transcript via the Grain MCP
- The MCP restructures the content around implementation's frame: technical requirements, integration questions, key stakeholders, workflow needs
- The structured brief is delivered to the implementation team
What it produces:
A clean implementation brief from a sales discovery call, organized around what the implementation team needs to do their job.
"Our sales team uses Grain exclusively," Jess said. "I take all of the information and I have it structured to what our implementation team cares about, and it processes all of that and feeds it back out. That's all the MCP — which is amazing."
Copy-paste this prompt to build it:
When a new sales discovery call transcript is available in Grain, pull the full transcript.
Restructure it into an implementation-ready brief, organized around what our implementation team needs rather than the order things were said in. Use these sections:
- Technical requirements mentioned
- Integration questions raised
- Key stakeholders and their roles
- Workflow needs specific to this customer
Send the brief to [Notion page / Slack channel / email] for the implementation team. Use plain language, and flag anything ambiguous instead of guessing at it.
Why this is worth copying:
The sales-to-implementation handoff is one of the most difficult processes in B2B software. Sales discovers things implementation never hears. This workflow requires zero behavior change from the sales rep.
Two workflows at a glance
The pattern across both workflows
Context, not content. The value in each workflow isn't the summary. It's the compiled context from dozens of conversations, structured to match what a specific reader needs right now.
No engineering required. Both workflows were built in plain language. No API configuration, no code. Jess built both workflows herself, without engineering support.
Behavior change is optional. Neither workflow asked sales reps or leadership to do anything differently.
The takeaway
Don't assume the Grain MCP can't do something. Describe what you want in plain language: what the output should look like, when it should run, who it's for. If the MCP can't do exactly that, it will suggest a workaround. The default assumption should be that it can do it.
What to try this week
New to Grain MCP? Start with the daily executive readout. Create a Notion database, connect Grain to Claude, and ask it to pull the last week of transcripts and surface the three things leadership should know.
Running a sales or CS team? Steal the sales-to-implementation handoff. Pull one recent discovery transcript and ask Claude to restructure it around what your implementation team cares about. Ship the manual version first, then automate it.
Not technical? Re-read the takeaway above, then tell the MCP exactly what you want. Start with one thing you wish you knew before every customer call.
Frequently asked questions
What are Grain MCP workflows?
Grain MCP workflows are automated processes that use the Grain Model Context Protocol and Claude to pull meeting data (transcripts, clips, summaries) into other tools on a schedule or trigger. You describe what you want in plain language; the MCP handles retrieval and Claude handles processing. No code required.
Do I need engineering support to build Grain MCP workflows?
No. The workflows in this post were built entirely in plain language, without custom code or API configuration. Jess Herbert built her executive readout and sales handoff workflows herself, alongside running GTM operations on a small lean team.
How do I build a workflow like Jess's?
Use the copy-paste prompts in this post as a starting point. Swap in your own Notion database, Slack channel, or email address, then adjust the sections to match what your team actually needs. Run it manually once before you automate it.
Is the Grain MCP available to all Grain users?
Yes, just open ChatGPT or Claude and find Grain. See our support page for current availability and setup details.
Resources
- Connect Grain to Claude: developers.grain.com/mcp
- Book time with the Grain team: grain.com/demo
