AI / ML
LLM applications, RAG, NLP, supervised learning, anomaly detection, clustering, neural networks, prompt engineering and model evaluation.
Software Engineer · Applied AI, Agentic Systems & Customer Solutions
Software engineer
building applied
AI products.
Software engineer with 7+ years of product experience, now focused on building applied LLM systems. Most recently an AI Engineer Intern at Narwhal, where I developed an agentic AI assistant for maritime logistics, while completing the University of Cambridge Data Science & AI programme.
01 / Skills
What I work with
Applied AI, ML and software engineering tools used to build and ship practical products.
Frameworks & Tools
Python, TensorFlow, scikit-learn, XGBoost, LangChain, OpenAI API, Streamlit, pandas, NumPy and Jupyter.
Engineering
JavaScript, TypeScript, React, Next.js, Node.js, SQL, API integration, Git, testing and accessibility.
02 / Professional experience
Where I’ve worked
Experience building software with technical and non-technical stakeholders.
AI Engineer Intern
The Narwhal Project
Architected and delivered an MVP AI assistant for maritime logistics, taking an ambiguous operational problem through design, staging release and customer UAT.
- Made key decisions across agent orchestration, APIs, database integration and workflow boundaries.
- Introduced Langfuse tracing and evaluation to investigate failures, grounding and response quality.
- Evaluated privacy and delivery trade-offs across AWS Bedrock and direct model inference.
Front-end Web Developer
National Foundation for Educational Research
Built and shipped accessible React and TypeScript assessment components, from requirements and feasibility through QA, validation and classroom testing.
- Developed 10+ production components tested with hundreds of primary-school students.
- Supported an AI-assisted marking pilot as the bridge between stakeholders and ML engineers.
Front-end Web Developer
Blue Fox Technology Ltd
Developed production GIS and mapping functionality for software used by councils and local authorities, including tools supporting planning-application workflows.
Freelance Web & WordPress Developer
Freelance
Delivered six end-to-end web projects for local businesses, owning requirements discovery, solution design, implementation and deployment.
03 / Projects
Selected work
Applied AI and machine-learning projects from my GitHub.
RAG / Retrieval↗
ML WikiTutor
End-to-end retrieval-augmented generation system over a curated ML/AI knowledge base of 23,124 chunks and 9+ million tokens, with source previews and refusal behaviour to reduce hallucination.
Conversational AI↗
LLM Chat for Kids
A child-friendly chat application with spoken responses, built around the OpenAI Responses API and a custom ElevenLabs character voice.
Machine learning↗
Student Dropout Prediction
Multi-stage supervised learning workflow predicting dropout risk across three stages of the student journey, framed around practical intervention timing and business-facing recommendations.
Anomaly detection↗
Ship Engine Anomalies
Unsupervised anomaly detection for ship engine sensor data in a predictive-maintenance context, comparing IQR, One-Class SVM, and Isolation Forest with PCA for visual analysis.
04 / Toolkit
The tools behind the work.
Applied AI & agents
Claude Agent SDK, Anthropic SDK, OpenAI API, LangChain, LangGraph, structured outputs, prompt design and agent orchestration.
Evaluation & retrieval
Langfuse, agent tracing, LLM evaluation, grounding, failure analysis, embeddings, vector search, Qdrant and RAG pipelines.
Software engineering
Python, TypeScript, JavaScript, SQL, Node.js, React, APIs, modular architecture, Git and CI/CD.
Machine learning & data
scikit-learn, XGBoost, TensorFlow, Keras, pandas, NumPy, NLP, forecasting and classification.