productonboardinggetting-startednew-customers

What to Expect in Your First Week with Stomme AI

You've signed up. Your agent is being configured. Here's what happens next — day by day, from the onboarding call to the moment your agent starts anticipating what you need.

You've signed up. Maybe you're excited. Maybe you're sceptical. Probably both.

Here's what actually happens between clicking "subscribe" and having an agent that knows how you work. No marketing spin — just the sequence, the awkward bits, and the point where it clicks.


Before Day 1: The onboarding call

Within 24 hours of subscribing, we schedule a call. It's usually 30–45 minutes. Here's what we cover:

What you need ready:

  • Your Mac (the one the agent will run on)
  • Telegram installed (this is how you'll talk to your agent)
  • A rough idea of what eats your time — we'll ask

What happens on the call:

  • We walk you through the installer. It takes about 5 minutes.
  • We connect your agent to Telegram so you can message it like a colleague.
  • We configure the basics: your timezone, your name, your communication preferences.
  • We talk about your workflow. What do your mornings look like? Where do you lose time? What's the first thing you'd hand off?

The call isn't a demo. It's configuration. When we hang up, your agent is live.

What it's not:

  • It's not a sales pitch (you already bought it)
  • It's not an exam (there are no wrong answers about your workflow)
  • It's not the last time you'll hear from us

Day 1: The quiet start

Your agent is running. It's doing... not much. And that's normal.

Day 1 is about learning, not doing. Your agent:

  • Connects to your email (you grant access during onboarding)
  • Starts reading your inbox — not replying, not sorting yet, just understanding the landscape
  • Connects to your calendar and maps your schedule
  • Sends you a test message on Telegram to make sure the connection works

Your first briefing arrives that evening or the next morning. It'll be basic:

Morning briefing — Wednesday

Today: 2 meetings (10:30, 14:00). 17 unread emails, 3 flagged as potentially important.

It's not impressive yet. That's the point. Your agent is being careful, not bold. It's learning what "important" means to you.

What you should do: Reply to the briefing. Even just "thanks" or "the 10:30 is a big one, prep me for that." Every interaction teaches your agent something.


Day 2: First real tasks

Now you can start asking for things.

Start simple. Don't ask your agent to reorganise your entire business. Ask it to:

  • "Summarise the 3 most important emails from yesterday"
  • "What's on my calendar this week?"
  • "Draft a reply to [person] saying I'll get back to them Friday"

Your agent will be slightly formal, slightly cautious. It doesn't know your voice yet. When it drafts something that doesn't sound like you, tell it:

"Good structure, but I'm more casual. Less 'please find attached,' more 'here's the thing.'"

It adjusts. Not perfectly on the first try, but it remembers.

The approval loop: Every external-facing draft comes to you on Telegram for approval before it's sent. You read it, tweak it if needed, and confirm. This never stops being true — your agent doesn't get to skip approval just because it's been around a while.


Day 3: The check-in

We send you a message. Not automated — a real person asking:

"How's it going? Is your agent picking up the right signals? Anything feel off?"

Common Day 3 feedback:

  • "The briefings are too long / too short" → we adjust the format
  • "It's flagging the wrong emails as important" → we tune the criteria
  • "I forget it's there" → we suggest a specific task to try

This is the most useful feedback you'll give us. Day 3 is where small friction becomes a habit if we don't catch it.


Days 4–5: Building rhythm

By now, your agent has seen a few days of your email, calendar, and communication style. Things start to get more useful:

Briefings improve. They include more context — not just "you have 3 meetings" but "you have 3 meetings, and Marcus hasn't replied to the thing you need before the 14:00."

Research gets faster. Your agent starts anticipating what sources you prefer. If you keep asking for competitor pricing, it learns to check pricing pages, not blog posts.

Drafts sound more like you. The formal edge softens. Your agent picks up your sentence length, your tendency to start with "So" or "Quick one —", your preference for bullet points over paragraphs.

What to try this week:

  • Forward a document and ask for a summary
  • Ask a research question you'd normally spend 20 minutes Googling
  • Say "remind me to follow up with [person] on Friday" and see what happens

Days 6–7: The click

Somewhere around the end of the first week, there's a moment. It's different for everyone:

  • You open Telegram in the morning and the briefing is already there — and it's useful
  • Your agent flags something you were about to forget
  • A draft arrives and you send it unchanged because it's exactly what you would have written
  • You realise you haven't opened your inbox directly in two days

That's the click. Not the agent being perfect — it won't be perfect for weeks yet. It's the agent being useful enough that you stop thinking about whether it's useful.


What to expect over the next month

Week 2: Your agent handles most routine email triage without you reviewing every decision. Briefings become something you rely on, not something you evaluate.

Week 3: You start giving it more complex tasks. Longer research. Multi-step follow-ups. "Keep an eye on this thread and tell me when they reply."

Week 4: You have a bad day and realise your agent quietly handled five things you would have forgotten. That's when it stops feeling like a tool and starts feeling like a colleague.


The honest bits

It won't be perfect on Day 1. Or Day 7. Your agent learns by doing and by your corrections. The customers who get the most value are the ones who talk to their agent like a new hire — patient, specific, willing to correct.

You'll second-guess it. You'll re-read drafts that are perfectly fine because you don't trust it yet. That's normal. It fades around Week 2 for most people.

Some things won't work immediately. Maybe your email provider has a quirk. Maybe your calendar setup is unusual. That's what the Day 3 check-in is for. We fix things fast.

Your data stays yours. Everything your agent learns about you — your preferences, your writing style, your priorities — lives on your Mac. If you cancel, it stays there. We don't learn from your patterns to improve other customers' agents.


Getting started

If you're reading this and you haven't signed up yet: Start here →

If you've signed up and you're waiting for your onboarding call: check your email. We've probably already reached out.

If you're in Week 1 right now and something feels off: reply to your agent, or reply to the Day 3 check-in. That's what we're here for.


Stomme AI agents run locally on your Mac. Your conversations, files, and agent memory stay on your device. Learn how we handle your data →

Ready to meet your agent?

Set up takes under an hour. No technical knowledge required.

Start for free