Maria-Elena Tzanev

Wilson - AI coworker in Slack that builds reports, work tools, + more

Wilson is the AI coworker who lives in your Slack and does the actual work, not advice, finished files. Ask in a thread and he builds the report, dashboard, deck, internal tool, or code, then drops it right back in the chat. He connects to your real everyday tools (Stripe, HubSpot, GitHub, Meta & Google Adsโ€ฆ), pulls the real numbers himself, asks before anything risky, and never trains on your data. Named after a certain volleyball. Far more useful. ๐Ÿ›Ÿ

Add a comment

Replies

Best
Maria-Elena Tzanev
Hey Product Hunt ๐Ÿงก We built Wilson because of a pattern we kept seeing: the best people on every team spend half their week on work that isn't really their work. "Quick" decks. "Just pull the numbers." A dashboard by EOD. A little internal tool nobody has time to build. Most AI tools will happily tell you how to do those things. Wilson just does them. He lives in your Slack, has his own secure cloud workspace where he writes and runs the code, connects to the tools you already use, and hands back the finished file: a board-ready deck, a live dashboard, a working internal app, a real pull request. A few things we cared about: ๐Ÿ“ˆ He pulls real data from your real tools, so you're not pasting essays of context every time ๐Ÿ™‹ He asks before anything risky; moving money, emailing a customer, pushing code ๐Ÿ›ก๏ธ He never trains on your data (SOC 2, HIPAA, GDPR via the Dynamiq OS he's built on) And yes, he's named after the volleyball in Cast Away. The lifesaver that gets you through. ๐Ÿ›Ÿ We'd love your honest feedback, your hardest "can he actually do this?" questions, and your favourite "things my boss thinks take 20 minutes" stories. We're around all day. We can't wait to hear what you think! Thank you! ๐Ÿ™
Abdullah Bin Asad

connecting to stripe, hubspot, github, and ad accounts while also writing and running its own code is a lot of trust in one coworker. the asks before anything risky part is doing a lot of heavy lifting there, worth knowing exactly what counts as risky before this touches real revenue data