AI & Backend Engineer
Python • FastAPI • RAG • Machine Learning Systems
London, UK · MSc Artificial Intelligence · 4+ years as Software Engineer
I build production-minded AI applications, asynchronous Python APIs and machine-learning pipelines.
My recent work includes FastAPI services for AI chat and analytics, RAG workflows, DynamoDB-backed systems, LLM evaluation and containerised deployment on AWS. I'm also experienced with the MERN stack (MongoDB, Express, React, Node.js). Previously, I developed React micro-frontends and Spring Boot APIs for enterprise retail systems used across more than 3,000 stores.
For my MSc Artificial Intelligence dissertation, I developed an explainable flight disruption prediction pipeline using large-scale aviation data, including 1.3 billion ADS-B records.
- Building reliable AI and backend systems with Python and FastAPI
- Interested in RAG, agent workflows, model evaluation and explainable ML
- Comfortable across APIs, data pipelines, databases and cloud deployment
- Based in London and open to AI, ML and Python backend opportunities
| Project | What it demonstrates |
|---|---|
| Flight Prediction Pipeline | Large-scale aviation data processing, feature engineering, XGBoost, SHAP and low-latency inference |
| Cyber Awareness Response Bot | FastAPI chatbot with trusted-source retrieval, citations, moderation, MongoDB and Slack/Teams/web interfaces |
| FastGraphAI | Full-stack LLM workflows using FastAPI, LangChain, LangGraph and Streamlit |
| FastAPI CRUD + WebSockets | Async MongoDB CRUD, Pydantic V2, authentication, WebSockets and automated tests |
| Real-Time Stock Tickers | Async FastAPI services, background tasks, WebSocket streaming and schema-generated models |
| E-commerce Application | Full-stack application development and frontend/backend integration |
- Processed data derived from 1.3B ADS-B records for flight disruption modelling
- Evaluated 637 model configurations across more than 393K flights
- Achieved 80.5% weighted F1 and 71.2% disrupted-flight recall
- Delivered approximately 14.8 ms median model inference
- Helped support enterprise software used across 3,000+ retail stores
Built FastAPI services, RAG workflows, analytics tooling and cloud deployment components for AI-enabled applications.
Developed React micro-frontends and Spring Boot APIs, improved service performance and supported enterprise retail systems used across more than 3,000 stores.
I'm interested in collaborating on applied AI, machine learning and production Python projects.