AI Daily Check-in App

In Episode 7 (August 8, 2025), Sam showed Daily Check-in, a web app that replaced our stand-up form. Each team submits what they finished, today's priorities and any blockers. Every morning, AI summarizes all of it, blockers first, and posts the summary to a Slack channel for Sam and Jordan. Sam isn't a developer. He built it in about two hours, spread over a few days between meetings, launched it on a Monday and shipped a v2 that same week.

Why we built it

For years our teams have filled out a quick five-minute digital stand-up with three questions: what did you complete yesterday, what are today's priorities, and is anything blocking you? It makes each team name its top priority, and it surfaces blockers first thing in the morning instead of in the next meeting. As teams grew, reading eight, nine or ten of these forms every day got hard.

So I said, why can't we just have AI summarize this and post to a Slack channel?

— Sam Nadler, Episode 7

The stack

• Bolt: Sam vibe coded the app and the form here.
• Supabase: the database, plus the edge functions that run the AI summary.
• GitHub: the cron job that triggers the morning summary.
• Slack: where the summary lands, and where each team's individual check-in still posts.

Built it in Bolt, and, yeah, we use Supabase for the database and for the edge functions, and I used GitHub for the cron job.

— Sam Nadler, Episode 7

How we built it, step by step

1. Rebuild the form in Bolt. Sam replicated the Google Form/Typeform stand-up: pick your team, recently completed, today's top priorities, blockers or concerns, and an optional energy level.
2. Store submissions in Supabase.
3. Write an edge function that summarizes everything with AI. The prompt tells it to put blockers first, since those matter most.
4. Schedule it with a GitHub cron job. The summary posts to a founders' Slack channel a little after 10 AM Central, 8 AM Pacific.
5. Keep the original posts. Each check-in still goes to that team's own Slack channel, so you can read the full version when needed.
6. Add history and metrics. Check-ins by date, completion rate and on-time rate per team, and team management for adding new teams.
7. Ship, then iterate. Launched Monday, v2 out within the week from team feedback.

How it turned out

Teams started using it right away. The day before recording, the sales team reported in their check-in that they had run out of credits on a tool. It showed up at the top of the AI summary, and Sam topped up the credits immediately instead of reading the form later in the day. Sam was modest about it: not a huge deal, but teams notice when you react faster to what's blocking them.

He also saw a loose correlation in the metrics: the most diligent teams fill it out on time, though "it's not perfectly correlated." Jordan's reaction at 07:20 was that Sam had built "a full stack web app" from scratch. Before AI, Sam could edit a few SQL functions and that was it.

And, you know, now I'm learning how to build edge functions and cron jobs, and it's all pretty easy with AI. Cumulatively, probably took me two hours total.

— Sam Nadler, Episode 7

What we'd do differently

• Add permissions before people share sensitive things. Some teams handle sensitive information, and for now Sam asked them to leave it out until permissions are in.
• Have AI look back a week. Jordan's idea at 09:50: compare the last seven days of check-ins to see whether stated priorities actually get done, and catch anything that fell through.
• Consider folding it into existing tools. Sam said some of our other tools might be a better home for it.

For another internal tool on Bolt and Supabase, see Mate, and for how Bolt compares with other builders, Bolt vs Claude Code. More in AI in productivity.

FAQ

How do you build an AI daily standup summary for Slack?

Sam rebuilt the stand-up form in Bolt, stored answers in Supabase, wrote a Supabase edge function that summarizes all check-ins with AI (blockers first), and used a GitHub cron job to post the summary to a Slack channel each morning.

How long does it take to build a daily check-in app with AI?

About two hours in total for Sam, spread across a few days between meetings. He launched it on a Monday and shipped a v2 from team feedback the same week.

Can a non-developer build a full-stack app with Bolt and Supabase?

Sam did. Before AI he knew a little SQL and wasn't a developer; with AI he built a full-stack app with a database, edge functions and a cron job.

What should a daily check-in form ask?

Ours asks three things: what you completed yesterday, your priorities today, and any immediate blockers. There's also an optional team energy level.