[Note #1] How to Launch an Internal AI Lab and not die trying

Good morning {{first_name | operator}} —

Welcome to the first-ever issue of Operator Notes—a field guide for COOs, VPs, and scale-up operators navigating the AI shift without the hype. I’m Marc, and after 11 years inside a 1,200-person scale-up (from intern to COO), I’ve seen how messy, promising, and often misunderstood AI adoption can be.

This newsletter is my way of documenting what actually works—from frameworks we use inside the org to playbooks that help teams move faster, not just talk smarter.

What’s inside this note:

AI Ops Radar – Major AI-Driven White-Collar Job Cuts predicted
Field Guide – How to Launch an Internal AI Lab: success isn’t about use cases - it’s about ownership
Tool Test-Drive – Gamma.app (AI presentation creator) verdict

Let’s dive in. ☕️

📡 AI Ops Radar

Here’s what happened in AI this week—and why your ops team should care (plus the one move to act on it).

Headline – Microsoft, Amazon, and Ford Predict Major AI-Driven White-Collar Job Cuts

Senior executives from leading companies warn that AI could replace millions of white-collar jobs, with Microsoft already announcing thousands of layoffs directly linked to AI automation. The trend is accelerating digital transformation but fueling workforce anxiety.
Proactively managing workforce transitions, investing in upskilling, and communicating transparently about AI’s impact on roles and responsibilities will become key for ops leaders

Next action → Launch upskilling and reskilling programs; conduct a workforce impact assessment to identify roles most at risk and plan redeployment strategies

🗒️ Field Guide — How to Launch an Internal AI Lab

Where to house it, who drives it, and how to run experiments without chaos

Rolling out AI org-wide sounds great on paper—until you try doing it. With 25+ departments and no clear playbook, we needed a lightweight but structured way to discover, test, and scale ideas. So we launched an Internal AI Lab—not as a physical space, but as a program with three key pillars:

1. Central ownership, distributed execution

2. A repeatable discovery and prioritisation process

3. A bias for piloting small, scalable wins

Here’s how we approached it.

How We Set It Up

I run the lab from the business side, not engineering. That was intentional—AI rollouts live or die by context, not code. To support it, I brought in:

A Tech Lead to evaluate feasibility and tooling
A Systems Engineer to move from idea to MVP quickly

Each department nominated an AI Champion—a context-rich, forward-leaning operator who helps surface opportunities and drive pilots within their team.

Discovery: Where Do We Start?

We work through one department per week in a 60-minute discovery session:

What’s repetitive?
What’s high-volume?
What’s painful or slow?

We log every idea in a shared board and classify it using Gartner’s AI Opportunity Radar:

Everyday AI: Back Office (Finance, HR, Legal, IT)
Customer-Facing AI: Sales, CX, Marketing
Game-Changing AI: Product, Core Ops, R&D

We also use an internal AI Opportunity Cheatsheet to help teams clarify, cluster, and tag opportunities.

Prioritisation Without Chaos

In weeks 3–4 of the lab cycle, we evaluate each idea based on:

Potential value
Ease of implementation
Risk profile
Departmental readiness

From there, we greenlight 2–3 department pilots. Enough to make visible progress without overwhelming the org.

Key Takeaway

Don’t build a task force. Start small, stay nimble, and use AI Champions to decentralise discovery and ownership.

One Thing to Remember

AI success isn’t about use cases—it’s about ownership. If no one’s responsible for shipping it, it won’t ship.

🔨 Tool Test-Drive — Gamma (AI Presentation Generator)

Check it out at gamma.app

Verdict: ⭐️⭐️⭐️ / 5 WAIT
Why it’s great: Great for generating presentations out of text documents (hint, ask Chatgpt to give you the content for each slide

Why wait: Not quite ready for company presentations (unless you don’t care about using internal branding). Missing the option to load your own template and just fill out the content.

Next up

Next week we’ll cover how to kickstart AI Fluency Across Your Ops Team - learn all the tools, courses and processes we implemented to upskill 1000+ employees.

Got questions or a win to share? Hit reply — I read every email.

— Marc