The Lab Notebook

AI, made digestible.

Field notes from the AmPhi Labs Innovation Studio: the real technical depth in plain language, the way we actually build, and how to run leaner, greener operations.

Track 01

AI, Demystified

Machine learning, clustering, neural networks, tokens, computer vision, minus the jargon.

Track 02

How We Build

Design thinking and agile sprints. Fuzzy idea to working pilot in weeks, not quarters.

Track 03

Lean & Green Ops

The AI Waste Audit, sustainable packaging, cutting cost and material at once.

The series
0
Start Here

Where to start with AI: a practical guide

The sequence decides adoption and impact. Get the problem, the owner and the evidence in the right order, and the right technology selects itself.

Available now · 10 min

1
φ
Track 01

Beyond GenAI: AI's big picture

The umbrella in plain language: what ML is and its three styles: supervised, unsupervised and reinforcement learning.

Available now · 6 min

2
Track 03

Dynamic pricing: the same maths can serve profit and planet

The same maths priced you out of the World Cup final, and saves 157M meals a year. The goal is the strategy.

Available now · 6 min

4
Track 02 · Coming soon

Agile innovation: ship small, learn fast

One big bet, converted into small cheap experiments. The idea under the sticky notes.

Coming soon

5
Track 01 · Coming soon

Let's talk tokens

What a token is, why your bill counts them, and seven ways to use fewer.

Coming soon

6
Track 02 · Coming soon

Design thinking: fall in love with the problem

Empathy, reframing and fast prototypes: the difference between AI people use and AI people route around.

Coming soon

7
Start Here · Coming soon

Buy or build AI?

The question every company gets wrong first, and the third option that usually wins.

Coming soon

8
Track 02 · Coming soon

Rapid prototyping & lean startup

MVP doesn't mean shoddy. Why a working prototype beats a business case, every time.

Coming soon

9
Track 02 · Coming soon

What is a hackathon? And does it actually work?

Why 48 hours of constraint beats six months of committee, and what the evidence says.

Coming soon

10
Track 03 · Coming soon

Gemini vs. Anthropic: a computer-vision showdown

Same images, same rubric, two frontier models, for sustainable packaging.

Coming soon

The rest of the series is still being written. One piece at a time, as each is ready.

Get the Lab Notebook on LinkedIn

AI Worth Building is the newsletter edition, out every second Tuesday. Subscribe there for each piece, and follow the studio for what happens between issues.

Ready to see where AI actually pays off?

Reading is step one. An AI Value Scan maps and scores your highest-ROI AI opportunities on profit and planet, fully remote, in days not quarters, before you commit a euro to building.

Discover your ROI with an AI Value Scan →