AI Engineer · Python Backend · LLM Applications · Applied Machine Learning
London, UK · MSc Artificial Intelligence graduate
LinkedIn · Portfolio · Email
I build Python APIs, LLM applications and machine learning pipelines. My background spans enterprise software delivery at Infogain, AI chatbot backend development at Elemental Concept, and an MSc in Artificial Intelligence from Northumbria University.
At Elemental Concept, my work covered FastAPI services, conversation and session management, usage and cost analytics, document retrieval and response grounding, TruLens evaluation, and deployment validation for Docker services on AWS ECS. Earlier, I developed React micro-frontends integrated with Spring Boot APIs and supported application and infrastructure issues across more than 3,000 retail systems.
I'm seeking AI Engineer and related Python backend or applied ML roles.
| Project | Engineering focus | Explore |
|---|---|---|
| FastGraphAI | Compare LangChain calls and LangGraph workflows through a FastAPI backend and Streamlit chat interface. | Overview · Backend · Frontend |
| CyberChatBot | Cybersecurity awareness Q&A with trusted-domain retrieval, Hugging Face model integration, citations, input filters, MongoDB training state and Slack/Teams/web interfaces. | Code and documentation |
| Flight Disruption Prediction | Schedule-aware ADS-B feature pipelines, model comparison, data quality checks, SHAP explanations and Streamlit exploration. | Implementation on v3 |
For my MSc flight disruption project, I processed data derived from 1.3 billion ADS-B records, engineered flight-level features and compared Logistic Regression, Random Forest and XGBoost using chronological train/test splits. The selected model achieved 71.2% disrupted-flight recall, 55.0% precision and 80.5% weighted F1 across the reported evaluation. The project included a Streamlit dashboard and SHAP explanations.
These are results from my study dataset and experimental setup; see the project documentation and notebooks for the pipeline.
| Area | Technologies and methods |
|---|---|
| LLM applications | Retrieval-Augmented Generation (RAG), Generative AI, prompt engineering, LangChain, LangGraph, LLM APIs, Hugging Face, TruLens, response grounding and evaluation |
| Applied machine learning | Python, pandas, NumPy, scikit-learn, XGBoost, feature engineering, model validation, chronological evaluation, SHAP, exploratory data analysis and Streamlit |
| Backend and data | FastAPI, Pydantic, SQLAlchemy, REST APIs, WebSockets, async database patterns, DynamoDB, MongoDB, SQL and ETL/data pipelines |
| Cloud and delivery | AWS, Amazon ECS, Docker, Azure, Git, GitHub Actions, CI/CD, Linux, deployment validation and log monitoring |
| Earlier software engineering | React, TypeScript, JavaScript, Redux, Node.js, Express, Java/Spring Boot API integration, Swagger/OpenAPI and production support |
| Academic foundations | Machine learning, deep learning, neural networks, natural language processing and transformers |
- FastAPI CRUD application — API development, validation and application structure.
- FastAPI WebSocket stock tickers — real-time API and WebSocket patterns.
- SQLAlchemy ORM template — reusable Python database patterns.
- E-commerce application — earlier JavaScript application work.
My older repositories preserve coursework, algorithms, design patterns and frontend projects alongside my current AI and backend focus.