Skills & expertise
Full-stack and AI development
I work across the whole stack: React interfaces, Python APIs with FastAPI and Django, and LLM systems built with LangChain and LangGraph and monitored with LangSmith and Langfuse.
Frontend development
I build responsive, accessible user interfaces with React and modern JavaScript, on a solid foundation of semantic HTML and well-structured CSS. The goal is always the same: interfaces that feel fast, work on every screen size and are easy for a team to maintain.
- Component-driven React applications with clear state management
- Responsive layouts that work from phones to wide screens
- Accessible markup: keyboard support, sensible headings, good contrast
- Fast pages: lean bundles, optimised assets, good Core Web Vitals
Backend & API development
On the server I work in Python, building REST APIs and services with FastAPI and Django. I focus on clean data models, secure authentication, predictable error handling and tests that make changes safe.
- High-performance async APIs with FastAPI
- Full-featured applications and admin tooling with Django
- Authentication, permissions and multi-tenant data isolation
- Background jobs, integrations and well-documented endpoints
AI & LLM engineering
I build applications on top of large language models: retrieval-augmented generation (RAG) over your own data, tool-using AI agents and multi-step workflows. LangChain provides the building blocks; LangGraph adds stateful, controllable agent flows that behave predictably in production.
- RAG pipelines: chunking, embeddings, retrieval and grounded answers
- Agents that call tools and APIs safely, with clear guardrails
- Multi-step, stateful workflows with LangGraph
- Prompt design, structured outputs and model selection for cost and quality
LLM observability & evaluation
AI features need the same rigour as any other code. With LangSmith and Langfuse I add tracing, prompt management and evaluations, so you can see exactly what a model did, measure quality and cost, and improve it with evidence instead of guesswork.
- End-to-end tracing of every LLM call, tool call and retrieval step
- Evaluation datasets and automated scoring to catch regressions
- Token, latency and cost tracking per feature
- Prompt versioning and A/B comparisons
Quantum computing
I have a working understanding of quantum computing: qubits, superposition and entanglement, quantum gates and circuits, and algorithms such as Grover’s and Shor’s. I follow where quantum and classical computing can work together, including alongside AI.
- Core concepts: qubits, superposition, entanglement and measurement
- Quantum gates and circuit design
- Key algorithms and what problems they speed up
- Hybrid quantum–classical approaches
Need these skills on your team?
Whether it’s a new product, an AI feature or an API that needs to scale, let’s talk.