How a 4-Person Startup Saves Up to 40 Hours a Month with Grain's MCP

TL;DR

• Gary Sahota, co-founder of Tether (inventory management and demand planning software for consumer brands), uses Grain's MCP with Claude to automate sales proposals, action item tracking, and engineering ticket creation.
• He saves 30 minutes to an hour per deal across 40 deals monthly — roughly 20 to 40 hours a month, or one to two hours a day.
• His "proposal creator" Claude skill pulls conversation context from Grain and internal documents from Notion to auto-draft sales proposals.
• A daily action items skill asks Claude: "What meetings did I have today and what are the action items?" and pipes the answers into his Notion to-do list.
• After engineering and onboarding calls, a ticket creation workflow summarizes the meeting and drafts a Linear ticket for whatever was discussed.
• Gary's one-sentence summary of the MCP: "It is business context that is always there — but was very hard to synthesize before."

Why small teams can't afford to lose meeting context

A four-person company has no room for dropped context.

Gary Sahota runs sales and marketing at Tether, a software company building inventory management and demand planning tools for consumer brands. Three founders. One engineer. Gary runs 10 calls a day, manages 40 active deals at a time, and writes every proposal himself. Alongside Grain, he has connected Gmail and Google to Claude, giving his workflows access to email and calendar context on top of his call history.

Before Grain's MCP, the cost of that pace was invisible: notes after calls, scrambling to remember what a customer said three weeks ago, re-asking on the fourth call a question already answered on the first.

The MCP changed the math.

What is a Grain MCP sales workflow?

A Grain MCP sales workflow uses Grain's Model Context Protocol server to expose your recorded meetings — transcripts, summaries, action items — as queryable context for AI tools like Claude. Instead of rewatching recordings or digging through notes, you ask Claude a question and it finds the answer in your Grain history. For a sales team, that means your call archive becomes a living, searchable context layer.

Workflow 1: Gary's Sales Proposal Creator

What it is: A Claude skill that queries Grain for every recorded meeting with a prospect and combines that context with internal Notion documents to auto-draft a tailored sales proposal.

How it runs:

1. Gary ends a multi-call sales cycle (typically 3-4 calls over weeks)
2. The skill queries Grain MCP for all meetings with that prospect
3. It pulls relevant context from Notion — product details, pricing, standard terms
4. Claude synthesizes both into a first-draft proposal, specific to what the customer said they need
5. Gary reviews and sends‍

‍What it produces: A context-rich proposal draft in minutes, replacing 30+ minutes of manual note-pulling and writing.

I have a skill that I've created — a proposal creator — which will just basically create a proposal pulling the context of what we talked about from Grain, internal documents from Notion, and auto-draft a sales proposal that I can send off to somebody.

Workflow 2: Gary's Daily Action Items Sweep

What it is: A Claude skill that runs at the end of the workday — pulling action items from every meeting that day and writing them to a central Notion to-do list.

I could just have a Claude skill and say, 'Hey, what are all the meetings that I had today? And what are the action items of all my meetings and put them in my Notion to-do list?' And I'll just do all of that for me. So I can centralize my to-do list.

Workflow 3: Engineering Ticket Creation from Calls

What it is: After engineering or customer onboarding calls, the team uses the Grain MCP to summarize what was discussed and create a Linear ticket for any feature or bug that came up.

We'll talk about something live, use the MCP to summarize the conversation, and then create a Linear ticket for a new feature that we thought about in an engineering call.

The pattern across all three

• The value is in the retrieval, not the recording. Grain's MCP makes recordings queryable and composable without manual effort.
• Being present in the call is itself a return. "How important is it to be present and not have to be writing notes the whole time? That's the value of a call recorder."
• Small teams feel the ROI faster than large ones. No ops team to maintain the system. Gary built the skills himself.
• The limit is the token window. Bulk deal reviews across 20-30 hour-long calls hit limits. Workaround: run against Grain AI summaries, not raw transcripts

‍