MSc AI graduate specialising in LLMs, RAG systems, and high-throughput ML infrastructure. I build things that are fast, measurable, and grounded in research.
core stack
ML engineer at heart, researcher by training. I care about systems that are fast, explainable, and built to last.
I'm an AI/ML engineer with an MSc in Applied Artificial Intelligence from the University of Bradford. My work sits at the intersection of production systems engineering and applied research — I've spent the last two years building things that ship, not just things that benchmark well.
At Outlyst, I reduced call latency by 54% across 2,100+ concurrent AI voice interactions by profiling async I/O bottlenecks and restructuring connection pooling. For my dissertation, I built FinLaw-UK — a graph-augmented RAG system that improved legal-finance Q&A accuracy by 19% using Mistral 7B and Neo4j.
I'm interested in roles where I can continue pushing the boundary between research and production — whether that's LLM infrastructure, graph-based retrieval, or high-throughput ML systems.
University of Bradford
Dissertation: FinLaw-UK: A Graph-Augmented Retrieval Chatbot for Reliable UK Financial Regulation
Focus: Spatial and relational data modelling, graph networks, LLM evaluation, robustness benchmarking
COMSATS University Islamabad
Dissertation: AI-Assisted Analysis and Prediction of At-Risk Diabetic Individuals
Focus: Predictive analytics, interpretability, biological impact modelling
From research labs to production AI systems — a timeline of where I've built and what I've shipped.
Real problems, measurable outcomes, production results.
Graph-augmented RAG for UK financial regulation
Production-grade RAG system combining Mistral 7B with a Neo4j knowledge graph to deliver faithful, cited answers on UK financial regulation.
2,100+ concurrent AI sales calls at 1.1s latency
High-throughput outbound calling system with a 54% latency reduction achieved by profiling async I/O bottlenecks and restructuring connection pooling.
93% accurate clinical risk scoring with SHAP interpretability
Ensemble ML pipeline for diabetes risk prediction with SHAP-based per-prediction explanations and a production REST API — presented at ICSMAI 2024.
Unified job-search aggregator with ATS resume matching
Production fullstack platform aggregating 6 job boards in parallel with real-time SSE streaming, secure multi-user accounts, application tracking, and 100% client-side ATS scoring.
An AI assistant that knows about my experience, projects, and research. Ask it anything about my work.
AI agent ready. Ask me anything about Hammad's experience, skills, or projects. Try: "What's your strongest ML project?"
Powered by AI · Responses based on Hammad's CV
Peer-reviewed work presented at international venues.
Graph-augmented retrieval, LLM faithfulness evaluation, systems optimisation for high-throughput ML pipelines, and interpretable predictive modelling for clinical applications. Currently exploring MSCA-eligible opportunities — earliest start date October 2026.
Whether you have a research opportunity, an interesting engineering problem, or just want to say hello — I'd love to hear from you.