DevOps and Data Engineer | AI and LLM Research | MSc Distinction, UEA | Seeking PhD 2027/28
I spent three years as a DevOps and data engineer in FinTech, then moved into AI research. My research looks at when we can trust what LLMs produce, from the code they write to how they reason.
- Silent failures in LLM-generated CUDA kernels: When repair hides the bug (with Prof. Stephen Laycock). MSc dissertation, submitted to CIUK 2026.
- Reasoning divergence of LLMs on P-FOLIO (with Dr Farhana Liza). In preparation.
- Predictors of research output quality across academic disciplines (with Prof. Kenny Coventry). In preparation.
- RailSense: open-source rail AI platform with live rail APIs, delay prediction and a RAG assistant.
- Mathematical interpreter and compiler: interpreter, C transpiler and RISC-V compiler in F#. I built the AST, parser and type checker.
- Languages: Python, R, CUDA C++, C++, F#, SQL, Bash
- AI and data: LLMs, RAG, NLP, machine learning, PostgreSQL, Kafka, Debezium
- Infrastructure: Linux, Kubernetes, Docker, AWS, Terraform, Ansible, ArgoCD, GitLab CI/CD, Prometheus, Grafana