QC newsletter 41: I've been writing. Just not sending

Who's with me in trying to stay sane out there? I've been bouncing between embracing what life has to offer and the very strong temptation to hide under the duvet and pretend the world isn't happening. It's one reason why these newsletters have been sparse.

But I have been writing. Just not sending. This issue is a catch-up — a batch of things I've been working on and thinking about since we last spoke.

Be kind to yourself.

Anne-Marie

📘 Ask the Handbook

I built the Quality Coach Researcher thinking people would use it as a standalone tool — ask a coaching question, get an answer grounded in the Handbook. Useful, right?

Turns out I had the framing wrong. By analysing the usage, it turns out the people who got the most out of it weren't using it instead of the book. They were exploring a specific topic, seeing how deep the material went, and then making a purchase. It's a discovery tool, not a replacement.

So I've repositioned it. It now lives on the book page as Ask the Handbook — ask a question, get an answer in real time before you commit. Free, no sign-up. You can also switch roles — engineering manager, test lead, delivery lead — so answers are framed for your context, not just a generic quality coach.

[Try "Ask the Handbook" →]

Quality Coach's Handbook

My Genie Got It Wrong: Evaluating LLMs for a RAG Chatbot

When I built the Quality Coach Researcher, I assumed picking the right LLM would be straightforward. It wasn't. I ran 100 iterations comparing Llama 70B, 405B, and GPT-4, and the AI agent's own recommendation turned out to be wrong. This post is about what I learned and how I'd approach it differently.

[Read the post →]

🛣️ Tech Leadership

A distinctly human perspective on tech leadership.

Staying Sane in a Post Modern Tech World

2025 was genuinely a great year for me professionally — conferences, workshops, good work. And yet there's something disorienting about the pace of change in tech right now that I keep coming back to.

[Read the post →]

Not Technical Enough: Really?

I've had a shadow following me my entire career. Her name is "not technical enough." This post is a celebration of looking her in the face.

[Read the post →]

🌱 Around the Grounds

Having done my own evals, I find that what Katja Obring writes in this article makes a lot of sense.

The oracle has opinions - Kato Coaching

This is part of a series where I write about what I’m learning in an AI evals and analytics course. Earlier posts cover the basics of evals, sniff tests, and quantitative evaluation with LLM-as-a-judge. In traditional testing, the oracle is trustworthy. You know what the right answer is, or at least you know who decides. […]

Kato CoachingKatja Obring

Lisa Crispin is offering a free on-demand course on measuring quality improvement.

Continuous quality improvement supported by metrics - Holistic Testing with Lisa Crispin

Free online, on demand course to learn ways to set quality goals, design small experiments to work towards them, measure progress effectively

Holistic Testing with Lisa CrispinLisa Crispin

Supply chain attacks on npm packages and LiteLLM made dependency hygiene feel urgent. Marit van Dijk's post is a practical response — using IntelliJ IDEA to find out not just which dependencies are vulnerable, but whether your code is actually hitting the vulnerable API.

Read the post ->

Royalee Martin, Veronika Pliusnina and Alessandra Moreira have joined forces on the engineering quality podcast. Worth checking out this podcast for its diverse content. Fiona Charles and I were guests on the topic of Quality Coaching.

📅 Upcoming

Testing Talks Sydney — 6 August 2026 I'll be at Testing Talks again this year. If you're in Sydney, come say hello.

[Event details →]

Happy reading, Anne-Marie