Hey HN! here's something I couldn't stop thinking about:
your coworkers probably think good things about you that they'll never say. Not because they don't mean it, but because there's no moment in the workday where it's natural to say "you're really good at keeping things calm when everything's on fire."
Most recognition tools try to solve this by asking people to write compliments. That doesn't work. It's awkward, it's performative, and it ends up in a public channel where "Great job Sarah! " means nothing. The insight I kept coming back to is: what if you reframe the compliment as a poll answer? You're not choosing to compliment someone. You're just answering a question. That removes the awkwardness entirely.
So I built KudoSnap. It's a Slack bot. Here's how it works:
1. One person installs it — no admin approval needed. The Slack scopes are intentionally minimal (channels:read, users:read, chat:write). No message access.
2. It builds a social graph from channel co-membership to figure out who actually works together.
3. A couple times a week it sends you a 3-second anonymous poll: "Who explains complex things the clearest?" with 4 names from your actual coworkers.
4. Your vote becomes a private DM to that person: "A coworker said you explain complex things clearly." They answer a poll to see who said it — and now they've also voted for someone. The loop spreads itself.
The whole thing runs on Vercel + Supabase ($0/month at beta scale). pg_cron triggers poll delivery via HTTP calls to serverless endpoints.
It's Employee Appreciation Day, which felt like the right moment to share, but the reason I built this is that appreciation shouldn't need a holiday.
Free beta, open to any Slack workspace. I'd love feedback, especially: what would make you skeptical about installing this?
alessandroetc•12h ago
your coworkers probably think good things about you that they'll never say. Not because they don't mean it, but because there's no moment in the workday where it's natural to say "you're really good at keeping things calm when everything's on fire."
Most recognition tools try to solve this by asking people to write compliments. That doesn't work. It's awkward, it's performative, and it ends up in a public channel where "Great job Sarah! " means nothing. The insight I kept coming back to is: what if you reframe the compliment as a poll answer? You're not choosing to compliment someone. You're just answering a question. That removes the awkwardness entirely.
So I built KudoSnap. It's a Slack bot. Here's how it works:
1. One person installs it — no admin approval needed. The Slack scopes are intentionally minimal (channels:read, users:read, chat:write). No message access. 2. It builds a social graph from channel co-membership to figure out who actually works together. 3. A couple times a week it sends you a 3-second anonymous poll: "Who explains complex things the clearest?" with 4 names from your actual coworkers. 4. Your vote becomes a private DM to that person: "A coworker said you explain complex things clearly." They answer a poll to see who said it — and now they've also voted for someone. The loop spreads itself.
The whole thing runs on Vercel + Supabase ($0/month at beta scale). pg_cron triggers poll delivery via HTTP calls to serverless endpoints.
It's Employee Appreciation Day, which felt like the right moment to share, but the reason I built this is that appreciation shouldn't need a holiday.
Free beta, open to any Slack workspace. I'd love feedback, especially: what would make you skeptical about installing this?