
Helly is a proactive AI agent that understands your work across tools-- and suggests actions automatically, without waiting for prompts.
Helly is a proactive AI agent that understands your work across tools-- and suggests actions automatically, without waiting for prompts.
Amazing experience! Incredibly useful addition to my toolkit. It already eliminates the majority of my tedious workflows.
So much of my day is consumed by meaningless messages and admin work on Jira. Helly is changing the game!
Reel.fyi
Hey Product Hunt — Dake here 👋
Why I’m building Helly 🧠
While working in product management at Microsoft, Walmart Search, and some scrappy startups, my day was constantly shredded by one-off, context-heavy asks—someone pings for an update in Slack, I dig through five tabs to cross-check Jira, Notion, and a spec, and my focus is gone. Every switch steals momentum. I needed an assistant that notices the request, pulls the pieces together, and nudges the next step without me stopping to prompt it.
What Helly does today ⚙️
Slack-first brains – reads messages, spots action items, and can fire them off for you (create the Jira ticket, send the reply, update Notion).
Browser record – grabs your cross-tab activity and shows a one-click Run for any task you want finished.
Early integrations – connect your apps once; Slack, Jira, and Notion are live.
Where we’re headed 🚀
Global context across all your tools / apps – any window, any tool.
Knows your work – maps how files, tickets, and chats relate so answers are one hop away.
Get your focus back – those context-heavy one-off tasks run in the background, so you stay on track.
We’re opening early access to just 100 people to keep quality high. A good chunk of spots are already filled from early conversations, but there’s still space on the waitlist.
@dake_zhang1 Just hopped on the waitlist can’t wait to see how it navigates complex Slack threads.
How detailed is Helly’s control over task execution? For instance, if it updates Notion or creates a Jira issue, can users set constraints or review things before they go live?
Reel.fyi
@ruby_henry good question! You’ll get details of the exact tool calls (like creating a Jira ticket) being used and their parameters (title, description, issue type, etc). You can edit these details too, and will need to accept the suggestion —it won’t auto execute without your confirmation 🙂👍
Is it trained solely on internal data, or does it also use external models for its intelligence?