09/27/2026 updated
MH
100 % available
Senior AI & Full-Stack Engineer | LLMs, RAG, AI Agents | TypeScript, React, Node.js, Python, GCP/AWS
Poole, United Kingdom
Only remote
Self-taught engineer. A-Level Spanish (A*). 11 GCSEs incl. Maths (8), Computer Science (7)About me
Senior engineer with 9 years' experience shipping production software for brands like JD Sports and Fortnum & Mason, a UK fintech and a national restaurant chain. I build AI tools people actually use: LLM assistants, RAG and agentic workflows, plus the full-stack and cloud work around them.
AI AgentsRustRAGTailwind CSSOpenAIJavaScriptPHPApplication Programming Interfaces (APIs)Artificial IntelligenceAmazon Web ServicesAmazon CloudfrontGoogle BigQueryMobile Application DevelopmentSoftware as a ServiceCloud Engineering
AI & LLM Engineering
Production AI products and internal tools built with Claude (Anthropic) and OpenAI models: RAG pipelines, prompt engineering, model evaluation and LLM features shipped to real users. Examples: a company-wide AI assistant for a national restaurant chain, an AI pull request review bot, and Poppy AI (Claude + RAG) at Popup Pal.
Agentic Workflows & AI-First Development
Daily use of Claude Code, MCP (Model Context Protocol) integrations and multi-agent systems to automate operations and accelerate delivery, with a strong focus on reviewing and securing AI-generated code.
Full-Stack Development
TypeScript, JavaScript, React, Next.js, Remix, Node.js, Express.js, Python, Flask, PHP and Rust. Responsive, accessible UIs with Tailwind CSS, Chakra UI and Material UI.
Mobile App Development
iOS and Android apps with React Native (Expo) and Native Modules, including App Store, Play Store and TestFlight releases.
APIs & Integrations
REST, GraphQL, OpenAPI/Swagger and Connexion. Integrations with Stripe Connect, RevenueCat, Commercetools, Onestock, FedEx, Prismic and enterprise APIs.
Cloud-Native, Serverless & DevOps
GCP (Cloud Run, Cloud Functions, Cloud Build, BigQuery, Firebase, IAM, Identity Platform) and AWS (Lambda, S3, DynamoDB, CloudFront, EC2, SQS, EventBridge, Route53). Docker, Kubernetes, Terraform, Terragrunt, infrastructure as code and CI/CD pipelines on GitHub, GitLab and Bitbucket.
Databases
PostgreSQL, Cloud SQL, Neon, Prisma, Firestore, BigQuery, DynamoDB, MongoDB, Redis and Aurora.
Testing & Monitoring
Jest, React Testing Library, Cypress, Detox, Mocha and pytest. Near 100% test coverage on the JD Sports OMS rebuild. New Relic dashboards and alerting.
Domain Experience
Retail and e-commerce (JD Sports OMS, restaurant stock control, SportsShoes.com replatform, EPOS and loyalty platform live in 10 sites, POS terminals for global brands), fintech and events/ticketing (Popup Pal).
Analytics & SEO
Google Analytics, Google Tag Manager and Google Ads for data-driven decisions.
Leadership & Delivery
Architecture decisions, mentoring and training developers, close stakeholder collaboration and Agile delivery (Scrum, Kanban). Founder of a live SaaS product, run end to end.
Languages & Right to Work
Bilingual English and Spanish. British/Spanish dual nationality, able to work in the UK and EU.
Experience
Professional experience5 - 10 years
Freelancer since2018
Languages
EnglishNative speakerSpanishNative speaker
Project history
Senior AI Engineer at a UK fintech lender, building internal AI tools.
- Build AI-powered internal tools that streamline workflows, reduce manual effort and improve operational efficiency.
- Work closely with stakeholders to understand operational challenges and translate business needs into practical AI solutions.
- Keep delivery aligned with internal priorities through regular stakeholder communication, clear requirements and fast feedback loops.
- Develop solutions across a wide range of projects in PHP, Python, TypeScript and Rust, adapting quickly to different languages and codebases.
- Review and strengthen AI-assisted projects, identifying and fixing security concerns in rapidly generated code to keep internal tooling secure.
- Apply engineering judgement to AI-assisted development, balancing delivery speed with security, code quality and maintainability.
Tech: LLMs, AI-assisted development, PHP, Python, TypeScript, Rust
Front-end and Order Management System (OMS) rebuild for JD Sports, delivered through Cabiri.
- Played a key role in leading the front-end and OMS rebuild for JD Sports, improving user experience and operational efficiency.
- Architected and managed a complex NX monorepo, streamlining development across multiple packages and teams.
- Used AWS and Terraform to design, configure and deploy scalable, secure, highly available cloud environments for critical applications.
- Implemented testing across every layer of the application, reaching near 100% test coverage and significantly reducing production incidents.
- Led integrations with Commercetools and Onestock, strengthening JD Sports' e-commerce and inventory management capabilities.
- Designed and deployed New Relic dashboards and alerts for real-time monitoring and early detection of issues.
- Worked closely with business leaders and technical teams to choose solutions aligned with business goals and best practice.
- Ran in-depth training and knowledge transfer so the JD Sports team could maintain and extend the new system independently.
Tech: AWS, Terraform, NX monorepo, Commercetools, Onestock, New Relic, automated testing
Founder and lead engineer of Popup Pal (popuppal.com), a community-first events and ticketing platform live on iOS, Android and web.
- Took the product from concept to production solo, with live organisers, paid ticket sales and real events running on the platform.
- Architected and built the full stack: React Native (Expo) apps, Next.js web app, Express APIs with Prisma on Neon PostgreSQL, deployed to GCP Cloud Run, with Firebase Auth, Stripe Connect and RevenueCat.
- Built Poppy AI, the in-app assistant powered by Claude: writes event descriptions and marketing copy, powers personalised recommendations and answers support queries grounded in live platform data using RAG.
- Engineered an end-to-end QR ticketing and entry system, used live at real events with 38+ stallholders.
- Built an AI-driven operations layer: automated daily Slack digests covering Gmail, GCP, Jira, App Store Connect and customer activity, saving hours of admin every week.
- Pioneered an agentic, AI-first workflow with Claude Code, multi-agent setups and MCP integrations into Jira and Confluence, so a one-person team ships at the pace of a squad.
- Run the business end to end: store releases, UK trademark, Stripe onboarding, outsourced QA and organiser acquisition. Also delivered client websites under CODEHARDY.
Tech: Claude API, RAG, MCP, React Native, Expo, Next.js, Express, Prisma, PostgreSQL, GCP Cloud Run, Firebase, Stripe Connect, RevenueCat