I Asked My AI Chief of Staff What Else It Could Automate. It Found 12 Things.

Here's the full list, with what each one actually does and why the AI flagged it:

1. Auto-draft follow-up emails after meetings - Before: I'd finish a coaching call, jot notes in Granola, then forget to follow up for 3 days. After: the agent reads Granola transcripts, drafts a follow-up email within 30 minutes of the meeting ending, and puts it in my Gmail drafts. Estimated time saved: 15 minutes per meeting, 5-6 meetings per week.

2. Payment reminder drip sequences - Before: I manually checked which invoices were overdue and sent awkward reminder emails. After: the agent monitors Stripe and Xero, auto-sends a friendly nudge at 7 days, a firmer one at 14 days, and flags anything at 30+ days for my attention. Estimated time saved: 45 minutes per week.

3. Enrichment for Founder Signal signups - Before: I'd see a signup email and think "I should look them up" then never do it. After: the agent grabs the signup email, pulls their LinkedIn profile, startup details, and recent activity, then adds a task to Todoist with full context. This one changed how I onboard community members.

4. Stale thread detection - Before: email threads would quietly die because I forgot to respond. After: the agent scans for threads where someone replied to me 3+ days ago and I haven't responded, then surfaces them as a daily digest. Kills the "sorry for the late reply" problem.

5. Candidate application screening - Before: I'd manually review every application that came in for programmes or roles. After: the agent pre-scores applicants against criteria I define, adds a summary and recommendation, and only escalates the ones that need my actual judgement.

6. Sponsor response thread automation - Before: sponsor outreach emails would sit in my inbox for days while I context-switched between coaching calls. After: the agent drafts templated responses based on sponsor tier and inquiry type, ready for my quick review and send.

7. Meeting density monitor - Before: I'd book meetings until my calendar looked like a Tetris board and wonder why Tuesdays destroyed me. After: the agent alerts me when any day exceeds 4 meetings or when there's no buffer between back-to-back sessions. Simple circuit breaker.

8. Client relationship decay alert - Before: I'd lose touch with founders I was coaching because life got busy. After: the agent tracks when I last had meaningful contact with each active client and pings me if someone drops off for 2+ weeks. Prevents relationships from going cold.

9. AI audit prospect pipeline automation - Before: leads from the AI audit page would come in and I'd manually track them in a spreadsheet. After: the agent captures the lead, enriches the profile, scores urgency based on their message, and adds them to a pipeline in Todoist with next steps.

10. Content deadline trigger - Before: newsletter deadlines would sneak up on me because I was deep in product work. After: the agent monitors my publishing schedule and nudges me 48 hours before any content is due, with a draft outline pulled from my recent notes and conversations.

11. Business vs personal email classification - Before: personal admin (strata, insurance, medical) would get buried in my business inbox and I'd miss deadlines. After: the agent auto-tags personal emails and routes the tasks to a separate Todoist project so nothing falls through the cracks.

12. Personal admin task extraction - Before: personal emails would sit in my inbox as vague reminders. After: the agent identifies action items in personal emails, extracts them as structured tasks with due dates, and sends them to Todoist. Built this one live on stream in 15 minutes.

My first reaction? Honestly, not that impressed.

A couple of these I'd already built. Some felt generic. But then I got to #3: Enrichment for Founder Signal signups.

That one hit different.

Founder Signal is my community of 250+ founders. When someone signs up, I get an email.

And every single time, I think: "I should look this person up. See what they're building. Maybe reach out."

Every single time, I don't.

Because it's not urgent. Because I'm in the middle of something else. Because my inbox is a battlefield and that email is just another soldier.

But if it automatically went to my to-do list? With their LinkedIn, their startup, their context already pulled?

That would be perfect.

That's the kind of automation that doesn't just save time. It closes loops I've been leaving open for months.

One suggestion made me laugh: #7: Meeting density monitor.

The AI had flagged that I'd had 6 meetings in a single day.

Not some inhuman calendar nightmare. Just 6.

But 6 meetings on a Tuesday, back-to-back coaching sessions, no buffer time? That's my chaos threshold.

The AI knew this because it had access to my calendar. My Granola notes. My Bear notes where I'd written "Tuesdays are killing me."

It wasn't guessing. It was pattern-matching my actual life.

And it was right. Tuesdays are my density problem. I need a circuit breaker.

Of the 12 suggestions, I picked #12: Personal admin task extraction.

Why? Because it was bothering me right then.

I'd just received an email from my strata manager. Something about roof tiles. Completely unrelated to my business. But it needed to go on my to-do list before I forgot.

Usually, I'd copy-paste it manually. Or worse, leave it in my inbox and hope I remember.

So I built the automation. Live. On stream.

The rule: if an email is tagged as personal admin, extract the task, send it to Todoist, mark the email as done.

Took me 15 minutes. Now it runs forever.

That's the thing about automation. You build it once. It works a thousand times.

I still haven't built #3 (the Founder Signal enrichment).

Or #4 (stale thread detection).

Or #7 (meeting density alerts).

Not because they're bad ideas. Because I'm still closing other loops.

But the list is there. Waiting. Ready.

And whenever I have 20 minutes and think "what should I automate next?" I don't have to guess.

The AI already told me.

People ask me all the time: "Why don't you just hire someone?"

Because I don't want to.

I want to stay a one-person company. Every automation is one less reason to hire.

Every workflow I build compounds. It doesn't take holidays. It doesn't need onboarding. It doesn't have a bad day.

It just runs.

And the more I automate, the more leverage I have. The more I can focus on the things only I can do.

Leverage beats labour. Every time.

I've invested in over 230 startups. I've coached hundreds of founders. And the same problems come up again and again.

• Invoice tracking
• Client follow-ups
• Meeting prep
• Content scheduling
• Expense categorisation
• Relationship decay

Every founder needs these solved. But most founders don't have time to build custom automations.

So here's what I'm building: an AI OS for founders.

Plug-and-play agents. Pre-built workflows. Install them like apps.

Because if my AI chief of staff can find 12 automations for me, imagine what it could find for you.

Watch this space. Check out the AI OS for Business page on batko.ai. It's launching soon.

You already know some of the things you should automate.

But you don't know all of them.

Because some patterns are invisible until you measure them. Some inefficiencies are so normal you don't see them anymore.

That's where AI shines.

Give it access to your inbox. Your calendar. Your notes. Then ask:

"What am I missing? What could you automate that would actually help?"

You'll get some generic answers. Some you've already done.

But you'll get 1 or 2 that make you think: "Why haven't I fixed this yet?"

That's the list you want.

Then you build one. Then another. Then another.

And six months from now, you're running a completely different operation.

Build once. Automate forever.