8+ years shipping scalable web apps, SaaS products, and AI workflows. I work end-to-end — from spec and architecture to backend, frontend, and deployment — so you can go from idea to launch with one technical partner.
Most founders I work with land on one of these. If any sound familiar, we should talk.
I turn rough product concepts into clear specs, system architecture, and an MVP plan you can actually ship — without 6 weeks of slideware.
I refactor full-stack codebases, stabilise CI/CD, fix the data model, and bring the system back into a state where adding features is fast again.
Frontend, backend, database, cloud, auth, deploy, monitoring — one technical partner instead of coordinating three contractors who don’t talk to each other.
I use Claude, Cursor, and agent CLIs daily — but every output gets reviewed, tested, and architected by a human who’s been shipping for 8 years.
Eight stages. None of them get hand-waved away. AI helps; it doesn't replace any of them.
Understand the business goal, users, and constraints.
Translate into clear, developer-ready scope and tasks.
Data model, API boundaries, auth, infra plan.
Frontend with React / Next.js. Real interactions, not mockups.
Node.js / NestJS. APIs, jobs, integrations, data.
Jest, Cypress, integration, manual smoke. Real coverage.
Docker, AWS, CI/CD. Observability before launch.
Read logs, listen to users, improve what matters.
I use AI agents and CLI workflows for leverage — never as autopilot. Founders get faster delivery without losing the engineering quality that keeps a product alive in year two.
The AI handles the volume work — boilerplate, scaffolds, drafts, test stubs. I handle the decisions that matter: what to build, how to structure it, what to refuse to build, and when “the agent’s suggestion” is actually a maintenance bomb six months out.
Founders don’t need a pure prompter or a pure engineer. They need someone who can hold both.
Most engagements blend a few of these. Some are 4-week projects, some are 6-month embedded partnerships.
Idea → working product. 4–10 week sprints with weekly demos and a real launch at the end.
Multi-tenant, billing, auth, dashboards. The non-glamorous backbone that paid SaaS needs.
Drop into an existing team and ship vertical slices — frontend, API, data, deploy.
Data model, contracts, jobs, queues. Designed to scale before scale is a problem.
Agents, RAG, CLI automation, prompt pipelines — integrated into your real codebase, not a demo.
Internal tools, scripts, CLIs that make your team faster every week, not just once.
Turning rough product ideas into engineering scope, milestones, and honest estimates.
Things on fire? I read logs, fix the root cause, and write the runbook so it doesn't repeat.
A few open repos. Most client work is private — these are the projects with public code to look at.
Enterprise systems, SaaS products, and internal tools — owned end-to-end across frontend, backend, and infrastructure.
No percentage bars. Just tools I actually use to ship every week.
A few public repos around AI-assisted engineering, CLI tools, and developer workflows. Most client work lives in private repos — these are the ones you can read.
If you're a founder, startup, or team with a product idea, an AI workflow problem, or a build that's stalled — let's talk. I reply within a day, in your timezone.