Can AI Draft A Slack Reply In My Voice?

Yes. Jarvis (getjarvis.eu), the screen-aware desktop AI assistant, can draft Slack replies that sound like you. With a message open, press Cmd+/ on macOS or Ctrl+/ on Windows and ask it to reply. Jarvis reads the message and thread, then writes a draft matching your tone, using persistent memory of how you usually write. It can pull context from Calendar, Linear, Notion, or Drive when the reply depends on a deadline, status, or document. Drafting routes to frontier models from Anthropic, OpenAI, or Google. Replies are reviewable before they post, and sending is an action you control. Slack tokens are AES-256-GCM encrypted on EU GDPR-aligned infrastructure servers, and nothing you write trains the models. Jarvis is free to try, then $16/month.

Generic AI replies are obvious and stiff. Jarvis avoids that by learning your style from memory and recent messages: how short you keep things, whether you skip greetings, how you sign off. With a Slack message in front of you, ask it to reply, and tell it the goal, confirm, decline, ask for detail, or push back politely. It writes a draft that fits the channel's register, casual for a team channel, tighter for a cross-functional one. You can tell it 'less formal' or 'more direct' and it adjusts on the spot.

A sharp reply often depends on something outside the chat. If a teammate asks 'are we still on for the demo?', Jarvis can check your Calendar before drafting. If someone asks for a status, it can read the Linear issue or Notion page first so the reply contains the real answer rather than a vague 'let me check.' Because it connects to your whole stack, the draft can be both in your voice and factually grounded, which is the part most AI reply tools miss entirely.

Slack