What does AI-first engineering actually look like inside a real software organisation?

In this episode of Tech Paired, Michael Phair is joined by Allan Mayberry, Senior Director of Engineering at Planet DDS, to explore how AI is reshaping software development, engineering leadership, team design, hiring, security, and the future of technical careers.

Allan leads a 100+ person engineering organisation and is actively driving AI-first development practices across teams working with tools like Claude and Copilot. But this conversation goes far beyond hype. Allan shares a grounded, practical view of what is really changing, what still needs human judgement, and why AI should be treated more like a co-pilot than an autopilot.

We discuss how engineers are moving from writing every line of code to orchestrating systems, how AI is changing the role of senior engineers, and why the basics of software engineering still matter more than ever. Allan also explains why teams that tried AI six or twelve months ago may need to revisit it, because the tooling has evolved so quickly that old assumptions are already outdated.

A major theme of the episode is team design. Allan breaks down Planet DDS’s move towards smaller pod-based teams, designed to reduce context switching, improve focus, get engineers closer to customers, and create more autonomy. He shares why traditional agile structures can struggle when code is no longer the main bottleneck, and why product, engineering, QA and design need to work more closely around shared outcomes.

We also dig into governance, security, and risk. Allan talks openly about the dangers of blindly trusting AI-generated code, the importance of proper peer review, automated testing, change logs, compliance processes, and maintaining clear accountability. From prompt injection and AI-assisted cyber threats to deepfake candidates and vendor outages, this episode explores the new security landscape engineering leaders now have to navigate.

For junior engineers and graduates, Allan gives honest advice on what to focus on in an AI-first world: learn the fundamentals, understand why good engineering practices matter, find mentors, and use AI to accelerate learning rather than replace it. For leaders, his message is clear: carve out time to experiment, get hands-on with the tools, understand where they help, and build guardrails before gaps become risks.

This is a practical, honest and wide-ranging conversation for engineering leaders, software teams, founders, recruiters, product leaders, and anyone trying to understand what the future of software development really looks like.

In this episode, we cover:

• AI-first engineering in real teams
• Claude, Copilot and agent-based development
• Why AI is not a replacement for engineering judgement
• Engineers as orchestrators, not just coders
• Guardrails, testing and pull request reviews
• The future of junior software engineering roles
• Pod-based teams and reducing context switching
• Governance, compliance and AI risk
• Prompt injection, security threats and vendor risk
• Deepfake candidates and AI in hiring
• What leaders should do now to avoid falling behind
• Why the fundamentals still matter

Tech Paired is the podcast from Tech Pair, exploring the people, teams and technology shaping the future of work.