Tech & skills
The paper I fold with
The tools and techniques I reach for when taking an AI product from idea to production.
LLM training
Taking open models and making them good at one job, cheaply.
Agents & RAG
Multi-agent systems with retrieval, orchestration and tool calling.
Evals & alignment
Measuring quality and steering models toward what users prefer.
Backend
APIs and data layers that keep AI products fast and reliable.
Cloud & DevOps
Shipping and scaling: containers, pipelines and the edge.
Data science
Recommenders, forecasting and the statistics underneath them.
NLP
Language models before and after the LLM era.
Computer vision
Detection, reconstruction and image pipelines.
ML & DL foundations
The classics that still win more often than you'd think.
How I think
Research habits
- Meticulous, and I start from first principles on every problem.
- Able to analyse, document and write research papers to a high standard.
- I take complete ownership of the work, and team up when the problem needs it.