I get hands-on with systems that have stopped scaling.
Not a report handed off from the outside. I work directly in the codebase, alongside the people who own it, until it's simpler, safer to change, and cheaper to run.
Two kinds of engagement
Diagnostic / audit
A time-boxed look at a system that's stopped scaling: where the complexity is, what can be removed, and the shortest path back to fast, safe change. You get a clear picture and a plan.
Ongoing / embedded
Hands-on work alongside your team to actually make the changes: architecture, AI adoption, cost, and the engineering practices that make speed sustainable.
What I actually do
Simplify the architecture
Find what can be deleted before looking for what can be optimized. Size the system to its real constraints, not a bigger company's.
Rebuild the safe path
Real local dev environments, honest test coverage, dependable pipelines, and monitoring that tells the truth, so deploying stops being an event.
Adopt AI where it earns it
Spec-first workflows, multi-agent orchestration, and prompt evaluation, introduced with enough structure that speed doesn't cost comprehension.
Rebuild shared understanding
Work directly with the engineers who own the code so the system stays legible after I'm gone, not just faster while I'm there.
Advisory
I selectively advise founders, CTOs, and investors on software architecture, AI adoption, and technical risk. Every area below is something I've carried hands-on accountability for — in the codebase, with the team, at the systems level — not a framework I recommend from the outside.
Founder / CTO Advisory
- Architecture decisions made early that shape everything after — before the wrong call becomes load-bearing.
- When to rebuild, refactor, or delete, before headcount becomes the default answer to a complexity problem.
- Reading a system you inherited and finding the shortest path to moving quickly on it again.
Engineering Leadership Advisory
- Diagnosing why a team has slowed down — where complexity crept in and what it would take to remove it.
- Adopting AI-assisted development with enough structure that speed doesn't cost comprehension.
- Turning deploy-as-an-event into deploy-as-a-non-event: test coverage, pipelines, and monitoring that tell the truth.
Investor / PE Advisory
- Reading a CTO's roadmap and delivery numbers the way an investor needs to, independent of that team's own framing.
- Assessing whether an engineering organisation can carry the growth plan in the deck.
- Technical debt as a carrying cost with real numbers attached, not a vague concern.
Technical Due Diligence
- Architecture and scalability risk under the load the growth plan actually implies.
- Cloud economics and AI capability — distinguishing prototype-grade from production-grade.
- Engineering organisation design and delivery process assessed as a business risk, not just an engineering one.
Tell me what's stuck.
Start with a 30-minute call. No prep needed, just tell me what's going on and we'll figure out together whether this is a fit.
Book a 30-minute callPrefer email? Send an email, or connect on LinkedIn.