For a long time I was proud of how I lead: give people the vision, give them trust, get out of the way. Then I handed a big new area to someone brilliant, did exactly that, and watched them slowly fall apart.

This episode is about situational leadership: the four stages every person moves through on every task that matters, the four styles that match them, the two ways leaders get it wrong, and the one insight that changed how I see my team. It ends with a challenge you can do in under a minute while you're still listening.

The sponsor of this episode is Wires Uncrossed.

Andrea Magnorsky started out writing Delphi and C#, went off to build video games, fell for functional programming and F#, and somewhere along the way accidentally invented a practice for helping teams understand their own systems.

In this episode, we pick up the decision-making thread from our conversation with Ammar and pull it somewhere new. Andrea and I get into why teams quietly mean different things by the same word, how "rework is just the modelling you didn't do," and the difference between hat, haircut and tattoo decisions. We dig into the advice process, how anyone can make a decision as long as they seek advice, not permission, why ivory-tower architects make worse calls than the people building the system, and how to move a team from consensus-by-default toward transparent, decentralised decisions without losing control.

This episode is sponsored by Wires Uncrossed.

We call almost any group of people "a team." But is it? Ammar Sagban has spent his master's research chasing that exact question, and his answer starts with a small shift that changes everything: teaming isn't a noun; it's a verb. Something you do, on purpose.

In this episode, Ammar is just weeks away from defending his thesis, having done something I once did and then ran away from — gone back to academia mid-career. We talk about what that journey gave him, why teaming is a verb, and the question most of us skip in the rush to be efficient: does this decision even need to be made by the team?

In this episode:

- Why going back to study mid-career is worth doing every ten years or so

- Teaming as a verb and the test of whether your "team" is really a team

- The fishbowl: are our coaches, facilitators and buffer roles helping the team, or getting in the way?

- Team decision-making and why "how are we deciding this?" matters more than the technique

- The tool Ammar built from his research, and the conversation he's hoping to start

Resources & links:

- Ammar's decision tool: https://decision-making-influence-reserach.ansense.nz/invitation

- Creating Great Teams: How Self-Selection Lets People Excel by Sandy Mamoli & David Mole

- Connect with Ammar Sagban on LinkedIn: https://www.linkedin.com/in/asagban/

This episode is sponsored by Wires Uncrossed, lifting the standard of software engineering in New Zealand, Australia and beyond.

Looking for an executive role in New Zealand is nothing like looking for a senior one. The market is smaller, the rules are unwritten, and most of what decides outcomes happens in rooms you're not in.

In this solo episode, Jakub shares three big ideas that reframed how he thought about the exec search, the small things that quietly make or break processes, and what to remember when it doesn't go your way.

This episode is sponsored by Wires Uncrossed.

After three years, Joekub is back, and this time they brought a startup, an AI agent that writes project updates in five minutes, and some very uncomfortable questions about your career trajectory.

Jakub and Joe (collectively: Joekub) reunite on Tech Waka for a wide-ranging conversation about what's actually happening at the coalface of AI-powered delivery. Joe left agile coaching to join an AI startup and hasn't looked back. Jakub is steering product and technology through the same wave from the C-suite. Between them, they've got enough delivery scar tissue to know when something is new and different.

They cover why extreme programming turned out to be a prophecy rather than a methodology. Why the spec is more important than the code now. Why your AI agent might be sandbagging its own performance to protect team morale. And the question that no one in a sprint planning meeting wants to answer out loud: if the agent is outperforming the humans, who gets the bonus?

Funny, honest, and grounded in real experience, this is what it sounds like when two people who've been around long enough to have opinions stop being polite about AI and just say what they think.

This episode is sponsored by Wires Uncrossed.

JD Trask has been building global tech companies from Wellington since 2007, first Mindscape, then Raygun, now Autohive. When ChatGPT arrived, he didn't wait to see what happened. He shut down his company for a week, made every employee pass an engineering hiring test using AI, and started rebuilding his entire organisation around agents.

In this episode, JD shares what he actually saw during that week, why he put TVs around the office showing each employee's token usage, and why he thinks the "AI makes mistakes" argument says more about the person making it than about AI. He also shares his AI economic maturity model, his take on headcount as an ego metric, and why the most important thing a CEO can do right now is get their hands dirty.

Autohive: https://www.autohive.com/

Raygun: https://raygun.com/

This episode is sponsored by Wires Uncrossed - helping teams across New Zealand, Australia and beyond untangle software delivery and build systems that truly perform. wiresuncrossed.co.nz

Lauren Peate and her team at Multitudes spent 15 months researching how AI is actually impacting engineering teams, following 500+ developers, surveying hundreds of leaders, and conducting in-depth interviews.

What they found challenges the hype: engineers merged 27% more PRs after AI rollouts, but also did 20% more out-of-hours work. The biggest predictor of success wasn't the tool , but it was leadership.

In this episode, we dig into what separates successful AI rollouts from struggling ones, why peer-to-peer learning matters more than mandates, and what engineering leaders should actually be measuring.

Get access to Lauren's research here: https://www.multitudes.com/research

This episode is sponsored by Wires Uncrossed.

Daniel Walters is a consultant, CTO coach, and former CTO of Seek Asia - one of Southeast Asia's largest job platforms. He is the founder of Great CTO, where he works with technology executives across New Zealand and beyond. He is also the co-creator of an AI-assisted engineering course, now running its eighth cohort.

In this episode, we cover:

• What drew Daniel from the operator role into coaching and consulting
• How he defines the difference between coaching, mentoring, and consulting, and why it matters
• The most common themes that come up when CTOs seek coaching support
• Why experienced engineers often resist AI adoption, and how leaders can bridge that gap
• The origin of the AI-assisted engineering course and what they have learned building it
• Why a cohort-based, experiential format is the right container for this kind of learning

Links

• Daniel's website
• AI-assisted engineering course