Engineering Leaders Can Be the Most Powerful People Today
How the Landscape Is Changing with AI
First-level engineering leaders — Team Leads, Tech Leads, Engineering Managers — are having a rough time.
2024 was harsh. Organizations are flattening. The era of free money is over, and leaders got hit (I covered that here). Amazon openly admits they’re increasing the ratio of individual contributors to managers. And “transactional managers” — those who mostly schedule, allocate, and approve — are the ones most at risk.
2025 looks louder — with AI supposedly taking everyone’s job.
But the data tells a quieter story. Check layoffs.fyi: job cuts aren’t even close to the 2022–2023 post-COVID wave.
So maybe the threat isn’t AI replacing engineers.
Maybe it’s engineers who use AI replacing those who don’t.
That’s not new — it’s just the next iteration of our industry. 🤷
The New Leverage Point
But let’s explore how the landscape changes for us, low- and mid-level engineering leaders. It seems we sit at a unique crossroads.
Yes, the ones who only shuffle tasks will disappear.
But the ones who embrace AI agents, product operating models, and data-driven decision-making will become the most powerful people in the room.
Why? Because they can:
Deliver code again
Validate ideas faster than ever before
Bridge product and technology
Recognize what “good” looks like — in architecture, process, and product
1. You Can Deliver Code Again
At some point, every engineering leader stops coding — not by choice but by erosion.
You join more meetings, handle hiring, manage 1:1s.
Your team tells you not to take critical tasks because you’re “the bottleneck.”
And one day you realize it’s been months since you wrote a single line of code.
That was my story.
Until AI agents changed it.
After years of drift, I can now build complete apps again — even with just a few hours of deep focus per week. Copilot was a warm-up. AI agents are a comeback.
2. You Can Validate Ideas Faster
Engineering leaders live at the junction of product and tech — the interpreters between PMs, designers, and engineers.
Traditionally, the flow was:
Product → Specs → Backlog → Delivery.
Weeks pass before anyone knows if the idea works.
But most PMs start with hypotheses, not certainty. Requirements are fuzzy. Engineers get frustrated by undefined edge cases.
Vibe-coding platforms now let product teams prototype without engineering help — which helps, but also widens the gap.
You can do better.
With AI agents and a clean tech stack (a prerequisite), you can prototype within your stack in hours.
Imagine this:
A PM asks for a submission form but doesn’t know if it should be long, scrollable, or multi-step. You take your design system, micro-frontend, and hardcoded schema — and generate multiple interactive prototypes by the end of the day.
You just compressed discovery from weeks to hours.
3. You Can Bring the Data
Modern systems generate overwhelming telemetry — product analytics, infra metrics, business KPIs.
Grafana dashboards are easy.
Understanding what’s happening in the product isn’t.
As an engineering leader, you’re the only one who sees across layers — product metrics (Mixpanel, Amplitude), business data (internal data warehouses), infra signals (Datadog, Grafana), and logs.
Now, AI gives you superpowers.
With a few scripts, you can blend these sources, analyze logs, or generate dashboards-as-code — without waiting for DevOps.
You’re not just observing systems anymore.
You’re interpreting them.
4. You Know What “Good” Looks Like
Engineering leaders hold the big picture — architecture, testing, standards, product context.
AI-agentic work is shifting the game entirely: faster iteration, different risks, new responsibilities. It brings unique challenges that only you fully understand.
What are these?
Translate vibe-coded prototypes into production-ready architecture inside your tech stack
Optimize your design system, APIs, coding standards, and docs for agentic workflows and your internal vibe-coding methods
Tighten feedback loops with evals so AI can iterate faster
Keep what AI builds secure, stable, and scalable
You understand both the studio (discovery) and the factory floor (delivery).
You can make them work as one system — fast and clean.
If you look for a guide in the new AI-first reality, check my book: From Engineers to Operators - AI Strategy Workbook for Engineering Leaders
The Shift
Here’s my prediction: teams will shrink.
A PM, a designer, and 2–3 engineers can now achieve what once required ten.
But that makes the engineering leader more critical than ever — as the integrator of AI, product, and technology.
The manager of tomorrow isn’t a scheduler.
They’re a creator again.
