We've been developers for 25 years. CTOs and tech founders for 13. Last month, one of our fractional CTOs shipped a full product alone, part-time - using AI agents for what used to require 4–5 people.
We've hired hundreds of developers. Built teams from 0 to 50+. Run a freelancer network across 10 time zones. We've seen every era of software development - from writing assembly code to managing offshore teams to shipping with modern SaaS tooling.
And nothing has changed as fast as the last 18 months.
Here's what we're actually seeing - not as analysts, but as practitioners running Tech Due Diligences every month for VC and PE funds across Europe, and as builders actively using these tools ourselves.
When we audit a scale-up today and see a 10-person engineering team, our first question is no longer "are they senior enough?" It's: what are they actually doing that an AI agent couldn't?
Often, the honest answer is: less than half of it.
Code review, test writing, documentation, boilerplate generation, bug triage, PR descriptions, simple feature work - all of this is being absorbed by AI agents at a pace that most engineering managers haven't fully reckoned with yet.
A 3-person team shipping fast with AI agents is not a risk. A 15-person team without AI workflows might be. Burn rate is no longer a proxy for ambition.
One strong engineer who knows how to orchestrate AI agents ships faster than a team of five who don't. This is already true - not in two years. Now.
The engineers winning in this environment are the ones who have shifted their identity from "person who writes code" to "person who orchestrates systems that produce code." That's a fundamentally different skill set - and it's one most engineering teams haven't deliberately developed yet.
The tools that matter right now: Cursor for AI-native coding, Claude for complex reasoning and code generation, n8n for workflow automation, Supabasefor rapid backend scaffolding. These aren't toys - they're the production stack of the fastest-moving teams we work with.
In every Tech Due Diligence we run in 2026, we now explicitly assess AI adoption as a signal of engineering maturity. The questions we ask:
The answers reveal a lot. A team with 80% AI tool adoption and clear workflow patterns is a different asset than a team where two engineers use Copilot occasionally.
If you're a Series A or B founder, the question isn't "should we adopt AI tools?" It's "how far behind are we, and how fast can we close the gap?"
The companies we see outpacing rivals with 3x their headcount aren't doing anything magical. They've made a deliberate decision to invest in AI workflow design - and they've brought in someone to implement it, rather than hoping engineers figure it out organically.
The playbook is being rewritten in real time. The gap between AI-native and AI-resistant engineering teams is widening every quarter.
We help scale-ups build AI-native engineering organisations and advise investors on AI adoption as part of tech due diligence. Get in touch if you'd like to discuss what this means for your company or portfolio.