Six months on: did the gains hold?
14 Sep 2026
In March I wrote about what happened when we gave five teams Claude access. Honeymoon numbers are easy. Here's what it looks like once the novelty's worn off.
The stack: how our AI setup actually fits together
14 Sep 2026
A map, not a tutorial. What sits where in our internal AI infrastructure, what each layer does, and how the pieces connect.
What two years of AI transformation taught me about change
8 Sep 2026
Not a victory lap. What surprised me, what took longer than it should have, and why the teams I expected to adopt fastest didn't.
Transcription quality: the hidden variable in conversation AI
1 Sep 2026
Everything downstream depends on the transcript. Scoring, insights, coaching. Get that layer wrong and the clever stuff on top is confidently wrong too.
MCP servers: what they actually are and why I keep building them
27 Aug 2026
Plain English explanation of the Model Context Protocol, for the technical leader who keeps hearing the acronym and hasn't had time to dig in.
Building an AI-ready engineering team
19 Aug 2026
"AI-ready" doesn't mean everyone learns prompt engineering. Here's what it's actually meant for hiring, QA, and how the team is shaped, two years in.
Small automations beat big agentic builds
11 Aug 2026
Everyone's building autonomous agents. We've built dozens of small things that each do one job. I think we're getting more done.
From sampling to everything: how QA is changing
30 Jul 2026
For twenty years contact centre QA meant scoring a tiny sample and hoping it was representative. It never was. That's over.
I built our knowledge platform from the source code. Here's how.
21 Jul 2026
Our product documentation was patchy and out of date. The code wasn't. So I generated the docs from the code, built the pipeline that puts them everywhere they need to be, and set agents running daily to keep them true.
The CTO and Head of Engineering split, when it works
8 Jul 2026
Most writing about the CTO and Head of Engineering relationship is theoretical. Here's what ours actually looks like day to day.
We indexed every customer call. Here's what happened.
25 Jun 2026
Hundreds of recorded conversations, sitting in a tool most of the company couldn't access. So we made all of it queryable through Claude.
Chatbot QA: the thing nobody's doing properly yet
16 Jun 2026
Contact centres QA their voice calls and mostly ignore their chat channels. Chat is growing faster. That gap is going to hurt.
Enable, automate, multiply
4 Jun 2026
The sequence that worked when rolling AI out across five teams. Not a framework. Just the order things had to happen in.
What I actually do with Claude every day
26 May 2026
People ask what "staying on the tools" looks like in practice. Here's a normal week, specifically.
What support deflection actually looks like
13 May 2026
More than half our support conversations now resolve without a human. Here's how the bot got that good, and what the other half tells you about where people still matter.
How our SDR team got more time on the phone
30 Apr 2026
The first team we rolled AI out to wasn't engineering. It was sales development. Here's what changed, what nearly killed it, and what I'd tell an ops leader doing the same.
Build vs buy: how we actually decide
21 Apr 2026
Two customers asked about the same thing in the same month, unprompted. Here's how we work out whether to build it, buy it, or walk away.
The real cost of an LLM feature
9 Apr 2026
How much does each AI feature in your product cost per customer per month? If you can't answer that, you're not ready to scale it.
We gave 5 teams Claude access. Here's what happened.
31 Mar 2026
Engineering has been using it for 9 months. The wider business rollouts started 4 months ago. We built automations around their actual workflows and measured what changed.
I'm a CTO and I write code most days
28 Mar 2026
AI changed what staying technical means as a senior leader. I'm not shipping production code, but I'm closer to the technology than I've been in years.