Launched this week

AskDoozy

Launched this week

An AI coworker in Slack that does your real work

56 followers

An AI coworker in Slack that does your real work

56 followers

Visit website

Team collaboration software

•

AI Engineer

•

AI Chief Of Staff

Doozy sits in your Slack. Someone mentions it in a thread with a real task ("why was Acme charged twice?", "post signups and churn to #leadership every Monday", "export last month's refunds by reason") and it goes and does it in your actual database, Stripe, GitHub, Linear, PostHog, HubSpot or anything with an API. The answer comes back in the thread with the files and a receipt for every action it took. One price per workspace, everyone included, 14 days free, no card.

Free Options

Launch tags:Productivity•SaaS•Artificial Intelligence

Launch Team

AskDoozy

Maker

📌

Hi, I'm Shiv. I built this because of a thing that happened over and over at my job. Someone in support would get a complaint. "You charged me twice." They'd ask in Slack. An engineer would say "on it", open a laptop two hours later, run a query, paste a number. Meanwhile the customer waited and the engineer lost their afternoon. Multiply by every "can someone pull the MRR for last quarter" and every "how many of the churned accounts had opened a ticket". So I built a bot that could go run the query itself. The first iteration was very hacky, it could barely read files sent in the thread natively but the AI agent running the loop was smart enough to extract the file from slack search and then get the task done, this was a promising signal and Doozy is that, rebuilt so other companies can use it without me SSHing into anything. At Spur (spurnow.com), 13 people use it every day since August, about 2,000 requests a month. What it is: - You mention @Doozy in a Slack thread with a task. It works in your own systems and posts the answer, files and all, in that thread. - Each task runs in its own sandbox. Your API keys and database credentials never go into the sandbox, a gateway outside adds them to each call. The agent can't leak a key it never had. - Tools that can destroy things (refunds, deletes) are off until an owner turns them on. There's a receipt for every action and an audit log nobody can edit, me included. - It keeps notes about your business in plain Markdown you can read and fix. Skills are procedures you write yourself. - One price per workspace. No seats, no credits. Usage is an allowance; at the limit it pauses and asks, it never bills you something you didn't choose. 14 days free, no card, no engineer needed to set up. I'll be here all day. Ask me anything, especially the uncomfortable stuff about giving an AI production access.

Report

2d ago

Doneas

Does it create new environment I need to validate later ?

Report

2h ago

AskDoozy

Maker

@bengeekly Each task runs in a thorwaway sandbox which is heavily monitored and logged so you have full transparency on what the agent is doing, but the agent's scratchpads and artefacts are preserved so that whenever you follow up agent does not start fresh it has memory of all the things it did last time

Report

4m ago

credentials in a gateway the agent never sees is the right design. my own dev agents get read access only, on separate accounts, so destructive tools off by default fits. a read replica option for db would make yes easier. but in this case, the agent mostly for monitoring, getting RCAs etc only.

Report

20h ago

AskDoozy

Maker

@rajendra_prajapat That's great, how have you created your dev agent ? and why do you only use it for monitoring and RCA, why not get it to complete the loop as much as possible and you come in at the very last step to verify and complete the loop

Report

7m ago