Building AI systems that retrieve, reason, and ship to production.
AI/ML Engineer with 10+ years in software engineering and 3+ years specialising in Generative AI, RAG, and Agentic systems for Financial Services and Automotive. MSc Artificial Intelligence, University of Aberdeen.
From problem formulation to production deployment.
I design, build, and ship end-to-end AI systems for enterprise use — not just notebooks. That means owning a problem from framing and model selection through evaluation, fine-tuning, and deployment into live products used by real teams.
Most of my recent work sits at the intersection of retrieval and reasoning: RAG pipelines, agentic multi-agent workflows built with LangGraph, and knowledge-graph-backed retrieval on Neo4j, layered on classic NLP and transformer fine-tuning when a smaller, cheaper model is the better answer.
- Enterprise Generative AI applications
- Retrieval-Augmented Generation (RAG) systems
- Agentic AI & multi-agent workflows (LangGraph)
- LLM fine-tuning — LoRA, QLoRA, PEFT
- Semantic search & vector databases
- Graph RAG & knowledge graph solutions
- Intelligent document processing & enterprise chatbots
Systems shipped, not just prototyped.
Ten projects spanning generative AI, classification, agentic retrieval, and model optimisation.
Meeting intelligence platform
Built a summarisation tool on Azure OpenAI and LangChain that extracts key points directly from call transcripts, replacing manual note-taking for client-facing teams.
Multi-agent RAG & knowledge graph POCs
Designed LangGraph multi-agent workflows, Graph RAG on Neo4j, and multimodal PDF search with Qdrant/CLIP — proofs of concept that went on to shape the team's GenAI roadmap.
Text classification model optimisation
Evaluated and tuned ModernBERT, DeBERTa, and SetFit for production text classification, then re-architected the pipeline around OpenAI embeddings to cut training time without losing accuracy.
LLM fine-tuning & sentiment classification
Optimised Flan-T5 and Mixtral with PEFT/LoRA for critical NLP tasks, and built a multi-class sentiment tool on GPT-3.5 Turbo using few-shot prompt engineering.
Enterprise-ready RAG system
Designed a production-grade Retrieval-Augmented Generation system with agentic orchestration on top, built for reliable, traceable answers over enterprise document sets.
MCP-based agent integration
Built an AI project on the Model Context Protocol (MCP), connecting LLM agents to external tools and data sources through a standardised, interoperable interface.
Multi-agent system with LangChain & LangGraph
Orchestrated a set of specialised agents with LangChain and LangGraph to handle multi-step tasks that a single LLM call couldn't reliably complete end-to-end.
LLM fine-tuning with LoRA, QLoRA & PEFT
Adapted open-weight LLMs to domain-specific tasks using parameter-efficient fine-tuning — LoRA, QLoRA, and PEFT — to cut compute cost versus full fine-tuning.
Generative AI series on AWS
Built and deployed generative AI applications on AWS, covering model hosting, orchestration, and integration into cloud-native services.
Prompt engineering on Azure OpenAI
Designed and iterated prompting strategies on Azure OpenAI models to improve output reliability and task accuracy across a range of use cases.
10+ years, four companies, one thread: shipping software that works.
- Built a meeting-summarisation tool on Azure OpenAI + LangChain, securing 2 clients and lifting revenue 25%.
- Optimised Flan-T5 & Mixtral with PEFT/LoRA; improved BERT/SetFit classification accuracy from 76% to 85%.
- Re-architected classification around OpenAI embeddings, cutting training time 4x.
- Delivered a GPT-3.5 Turbo few-shot sentiment tool at 94% accuracy.
- Modernised legacy Loan Details & Check Hold applications, cutting runtime 30% and clearing 15 long-standing defects in one sprint.
- Built a credit-card recommendation engine from purchase patterns, improving decision accuracy 30%.
- Forecasted product success rates at 87% accuracy using customer sentiment & ML models.
- Built a user-action test framework and an internal tool auto-generating test case templates, cutting time 40%.
- Promoted to Senior Engineer within 12 months.
- Designed and automated a client billing system in C# and SQL; optimised stored procedures for 2x faster retrieval.
- Delivered billing automation for 3 new clients in a single month.