Available for new work
Senior full-stack engineer with 7+ years shipping production SaaS — AI agents and LLM workflows, the APIs behind them, and the interfaces that make them usable.
What I do
Most products stall at the seams between these. I work across all four, so the seams are designed rather than discovered.
React and Next.js applications where the hard part is the data — analytics dashboards, AI-generated content review, and admin surfaces that stay quick as the payload grows.
Python and Node services built to carry AI workloads: typed REST and GraphQL contracts, async processing and queues, hardened auth, and observability wired in from the first deploy.
Content and commerce platforms the owning team can actually run — WordPress and Shopify builds, headless architectures, and content models designed so routine changes never reach an engineer.
LLM workflows that behave predictably in production: prompt frameworks, agent design with tools and memory, retrieval over vector stores, and the evaluation loop that keeps quality measurable.
Selected work
A few projects that show the range — with the outcome, not just the stack.
Led end-to-end development of review-response agents on OpenAI and Claude, with tool integrations and memory for autonomous operation — and a prompt framework that cut model cost 18%.
Resilient integrations with Google Business Profile, Yelp and Facebook Reviews — webhook-driven where the platform allows it, scheduled where it doesn't, reconciled either way.
React and Next.js dashboards surfacing AI metrics and generated responses for client decision-making — increasing engagement 28% by making the model's work inspectable rather than opaque.
Experience
Jan 2024 — Apr 2026
Led end-to-end AI agent development for review-response workflows on OpenAI and Claude, integrating tools and memory for autonomous operation. Built the prompt-engineering framework that raised response reliability while cutting model cost by 18%, and kept integrations with Google Business Profile, Yelp and Facebook Reviews at 99.9% data-sync uptime.
Oct 2022 — Dec 2023
Built cloud-native Python and Node services exposing REST and GraphQL APIs for AI workloads and partner integrations. Dockerised CI/CD onto Kubernetes accelerated release cycles by 40%, and redesigned SQL/NoSQL data models doubled pipeline throughput. Mentored engineers and introduced the testing strategy.
Apr 2021 — Oct 2022
Architected backend microservices supporting AI workflows behind secure REST and GraphQL APIs, and built the React/Next.js dashboards that made AI metrics legible to clients — increasing engagement by 28%. Webhook-based review sync with retry logic held 99.8% data freshness.
Jun 2019 — Feb 2021
Full-stack SaaS development in Python and Node: scalable REST APIs over a service-oriented architecture, data pipelines across SQL and NoSQL, and authentication and authorisation across microservices. Observability and proactive incident response held 99.95% uptime.
Writing
Most confusion about RSC comes from thinking of it as rendering. It's closer to serialisation — and once you see it that way, the rules stop feeling arbitrary.
3 min read
Most teams try to cut their model bill by picking a cheaper model. In my experience the savings are almost never in the model — they're in the context you were sending it that it never needed.
3 min read
A CMS schema is a contract between engineers and editors. The difference is that when you get an API wrong, tests fail — when you get a schema wrong, people quietly stop using it.
3 min read
I'm open to senior engineering roles, contract work, and anything that puts LLMs in front of real users. Tell me what you're working on.