The Problem

Education nonprofits run tutoring and after-school programs on volunteer coordination, but program directors have zero visibility into who did what or who's stuck — and can't afford the tools bigger teams use to solve this.

What StandupSense Does

It reads what volunteer teachers are already saying in Slack and turns it into an automatic standup — no forms, no check-in bot pinging anyone. Volunteers just talk normally. StandupSense filters out the small talk, uses Groq AI to figure out what each person completed, what they're working on, and what's blocking them, then logs it to Notion and posts a clean summary back in the channel.

The impact is concrete: a coordinator who used to chase five volunteers individually, or build a tracking spreadsheet by hand, gets this automatically — turning hours of weekly admin work into zero.

The Hardest Part

Honestly, the AI extraction came together fairly quickly. What actually took real debugging was the MCP integration — I hit session errors, wrong tool names, and a validation error from wrapping the request payload incorrectly. I could've just called Notion's API directly and moved on, but I wanted the MCP layer to be real, not just present to check a box. I kept at it until I got an actual confirmed page-creation response back through the protocol, not a fallback.

Why It Matters

If this saves a program director even a few hours a week, it's done its job.

Built With

  • block-kit
  • groq-ai
  • javascript
  • llama-3.3-70b
  • mcp-integration
  • node-cron
  • node.js
  • notion-api
  • notion-mcp-server
  • slack-bolt
Share this project:

Updates

Submission history