We recently hosted our first Tech Leader Exchange at the OfferZen Cape Town office, where 30 engineering leaders came together to tackle one big question: How do you lead well in the age of AI?
The discussions were framed around two key themes from our Engineering Leadership Report:
- What does “great” look like in an AI-enabled engineering team?
- How do leaders navigate the rising pressure to adopt AI responsibly?
Here’s what leaders shared — the insights, disagreements, and moments of clarity.
Redefining “greatness” in the age of AI
Fundamentals still matter, but they show up differently
Leaders agreed that curiosity, ownership, product thinking, communication, and good judgment remain core to excellence.
But AI is changing how these attributes manifest:
- Juniors ship usable code faster, sometimes masking gaps in fundamentals.
- Seniors spend more time on architecture to keep pace with AI-accelerated delivery.
- Developers who understand the business context are disproportionately valuable.
- The performance gap between AI-fluent engineers and others is widening rapidly.
“AI changed the tools, not the fundamentals — but it changed the pace, and that’s changed how we recognise greatness. Clarity matters more than speed.”
Discernment and adaptability are the new differentiators
A phrase that came up repeatedly: discernment. Not just using AI, but knowing when and how.
Leaders described “great” engineers as those who:
- Break down problems before prompting AI
- Validate outputs thoroughly
- Can still code without AI when needed
- Understand long-term quality, not just speed
- Know when AI isn’t the right tool
AI is amplifying strengths, but it also exposes weaknesses. Some developers thrive; others struggle. What leaders are watching closely is adaptability: Who can adjust, learn, and evolve as the craft shifts?
“We’ve got brilliant developers who are suddenly struggling, because their strengths aren’t the ones AI amplifies.”
Product sense is now a superpower
A surprisingly strong theme from the discussions was that the engineers making the biggest leaps weren’t necessarily the strongest coders, they were the ones who understood the business context best.
Leaders described how developers who grasp things like pricing mechanics, financial flows, logistics, or customer behaviour:
- make better technical decisions,
- write more meaningful prompts,
- and deliver solutions that actually solve the right problem.
“When someone understands the domain deeply, AI just makes them more dangerous — in a good way.”
In an AI-enabled team, product sense isn’t a “nice to have” anymore. It’s one of the biggest multipliers of impact.
The takeaway
Greatness is still about fundamentals, but fundamentals aren’t enough on their own anymore. Teams need engineers who combine craft, product insight, and AI fluency with the judgment to balance speed and quality.
2. Navigating the pressure to adopt AI
Middle managers feel the squeeze the most
Executives want faster delivery and AI-enabled features. Teams want better tooling and experiment time. Customers increasingly expect smart interfaces or automation.
Sitting between all of that are engineering managers trying to keep morale up and delivery stable.
“Everyone wants AI adoption, but no one is giving us more time or people to make it happen.”
Pressure looks different depending on company size
Roundtable conversations echoed what we saw in our report data:
- Startups: AI adoption feels natural, efficiency means survival.
- Enterprises: Dedicated teams and budgets mean lower risk.
- Mid-sized companies: The hardest environment – delivery pressure plus transformation pressure, with neither resources nor specialists to support.
This is where leaders raised concerns around:
- shallow experiments,
- messy expectations,
- technical debt piling up,
- QA strain,
- and anxiety in their teams.
What’s actually working (and repeatable)
Across the roundtables, leaders shared the approaches that were helping them make AI adoption sustainable:
- Structured exploration, not ad-hoc usage: Short trials, AI Fridays, cross-functional experiment squads. The key is bounded time and clear purpose.
- Guardrails: Policies for data, better test coverage, and clear guidelines on where AI is allowed or disallowed.
- Enablement over enforcement: Teams respond better when given tools and space to explore, not mandates to adopt.
- Internal wins first: Many leaders shared internal automation examples that saved meaningful time without touching customer-facing code — low risk, high reward.
🌱 The takeaway
The pressure isn’t going anywhere. The leaders who handle it best are the ones who create clarity: why they’re adopting AI, where it fits, what good looks like, and how they’ll protect long-term quality.
Final reflection: Leading through a moving target
Across both themes, one message came through clearly: AI isn’t replacing developers, it’s amplifying the people who already have strong fundamentals and good judgment.
But for leaders, the work is getting harder. They’re navigating changing definitions of excellence, uneven adoption pressures, team anxiety, shifting expectations, and rapidly evolving tools — all while still delivering roadmap commitments.
The Tech Leader Exchange made one thing clear: Leadership in the age of AI is less about speed and more about clarity.
👉 Clarity on what greatness looks like.
👉 Clarity on where AI fits.
👉 Clarity on the skills teams need next.
We’ll be running more Exchanges in the coming months. If you'd like to join, sign up to our waitlist here to be the first to know.