Gavril reads and sends project updates, tracks task accountability and guards against scope creep — so your team builds, and clients never chase.
Gavril sits between your trackers and your clients — answering questions, chasing deadlines and writing the reports, so humans only step in when judgement is needed.
Answers on WhatsApp & Telegram with live task status — every reply validated before it sends.
Daily and weekly summaries, deadline reminders and overdue nudges, out before your standup.
Every request is checked against the signed scope. Out-of-scope asks get flagged, priced and pencilled into a future sprint — automatically.
Workload, performance and wellbeing per teammate — recalculated every six hours.
Works on top of the tracker you already have — Jira · Asana · Trello · GitHub · Monday · Azure DevOps · Google Sheets, and talks to clients on WhatsApp, Telegram and email.
What Gavril is, who it suits, and where it stops — in plain answers.
Gavril is an AI project manager for agencies and software teams. It runs on top of the tracker you already use — Jira, Asana, Trello, GitHub, Monday, Azure DevOps or Google Sheets — and does the coordination work around it: answering clients, chasing overdue tasks, joining meetings and filing the notes, checking requests against the agreed scope of work, and writing the status reports.
Those features summarize and draft inside their own tracker, for the people already logged into it. Gavril does the project manager’s outward work instead: it answers clients in their own chat (WhatsApp, Telegram, email), sits in the meeting, and chases the people who owe something. And it does that on top of whichever tracker a team already has, so nobody migrates to get it.
Teams delivering client work — agencies, studios and software teams — where the cost is not tracking the work but relaying it: status questions, change requests, meeting notes and follow-ups. On purely internal projects half of Gavril (the client-facing half) sits idle, and a tracker’s own AI is the cheaper answer.
No. Gavril reads and writes the board you already have through its API, and the tasks stay where the team works. It adds the project manager, not another place to look — if Gavril is switched off, the board is still the board.
Scope guard checks every client request against the signed scope of work before anything moves. In-scope asks become tasks; out-of-scope asks are flagged to the project manager and pencilled into a later sprint instead of quietly being built for free. It is the one thing an AI can do that a busy PM reliably cannot: read the contract every single time.
Every outgoing reply is validated before it sends. Anything the board cannot answer, and anything from a client who sounds unhappy, is held and handed to the team instead of guessed at. Gavril is built to be the one that answers at 9pm — never the one that invents an answer.
Yes. Gavril joins the Google Meet call, follows it, and answers out loud when someone addresses it by name. Afterwards the notes are written, the decisions recorded and the work that was agreed is already on the board. In-person meetings work the same way, recorded from a phone.
Gavril is opening in batches while it is invite-only. Leave an email on the waitlist at gavril.me and the team brings you in; setup is connecting the tracker you already use and the chat your clients already message you in.
Gavril is opening in batches. Leave your email and we'll bring you in.
The same voice that sits in your meetings. Ask what it does, or tell it how your projects run.
A short voice call, straight from your browser.