Giza plateau · c. 2560 BCE
Before the first stone, there is only the plan.
Scroll to build
01 / Vision
I think in systems.
Before anything is built, someone has to see the whole of it: the outcome, the parts it breaks into, the order they depend on, and the people and tools that must move together. That's where I start. I take a vague goal and turn it into a plan a team, or a set of AI agents, can actually execute.
- 01Envision the outcome
- 02Break it into components
- 03Map the dependencies
- 04Coordinate the whole
02 / Beneath the surface
Making complexity make sense.
I'm drawn to what's underneath: why a system behaves the way it does, where a bug really comes from, which part of a design is quietly carrying the weight. Complexity rarely disappears. But it can be understood, and once it's understood it can usually be simplified.
03 / The machinery
Human creativity. Machine acceleration.
The builders didn't lift stone with bare hands. Sledges, levers, ramps, ropes and wetted sand multiplied what people could do. AI is today's version of that leverage. It can carry a large share of implementation, fast. What it can't supply on its own is the reason to build, the context, the architecture, and the judgement to tell a good result from a plausible one.
The workflow, c. 2560 BCEThe workflow, today
- QuarryIntent
- HaulContext
- RampGenerate
- SetReview
- InspectIntegrate
The tools change. The shape of the work doesn't.
04 / The structure
There's always a better way to build.
Good engineering is mostly a series of good decisions. I prefer solutions with fewer moving parts, code that's easy to reason about, and architecture that makes the next change cheaper rather than riskier.
- IFewer moving parts. Simplicity is a feature, right up until it costs correctness.
- IICorrect, then clear, then fast. In that order, and fast only where it matters.
- IIIFind the cause. Symptoms come back. Causes, once fixed, don't.
- IVRemove unnecessary work. Then specialise, when real requirements earn it.
- VBuilt to change. Today's design should make tomorrow's change easy.
05 / The capstone
Good software is more than software.
A pyramid was never just a pile of stone. It was built for a purpose people believed in. Software is the same. The application that wins isn't the one with the most features or the cleverest architecture; it's the one that solves a real problem so well that people trust it, return to it and are glad to pay for it. Design, clarity and delight aren't decoration. They're part of the product.
Built to work. Designed to matter.
- ValueSolves a problem worth paying for.
- ExperienceFeels obvious, fast and considered.
- TrustBehaves the same way every time.
- ReturnPeople come back without being asked.
06 / The work
From solving problems to creating value.
Seven to eight years of programming, three-plus professionally: web applications, APIs, databases, system architecture, infrastructure and some game development. The stack changes with the problem. The approach doesn't.
What I bring
- Product & requirements
- Turning vague ideas into concrete scope, flows and priorities.
- Architecture & workflows
- Data flows, boundaries and interactions that stay easy to reason about.
- AI-directed development
- The context, constraints and review that make AI tools produce work worth keeping.
- Diagnosis
- Tracing bugs and bottlenecks to their actual cause.
- Full-stack delivery
- Frontend, backend, databases and Linux infrastructure, end to end.
Recent tools: Vue · Nuxt · TypeScript · Python · FastAPI · Node.js · PostgreSQL · C# · ASP.NET · Unity · Linux. Means, not identity.
07 / Beyond
Build beyond the limitations of yesterday.
I'm looking for AI-assisted development roles, freelance projects and collaborations on products worth building well. If you have a problem that deserves careful thinking, let's talk.