Services
What I can build for you.
Four service tracks where I’ve shipped measurable outcomes in production. The toolkit below is what powers them.
01
LLM engineering
Fine-tuning, prompt engineering, RAG pipelines and evaluation harnesses tuned for production traffic — including sub-8B on-device deployments where latency and privacy are non-negotiable.
02
ML systems & data
End-to-end PyTorch and scikit-learn models with the evaluation discipline that survives peer review: baselines, ablations, statistical significance, reproducibility scripts.
03
Edge & embedded AI
Edge Impulse workflows for Snapdragon X, Meteor Lake and Hailo — small-footprint models shipped where the network is unreliable and the round-trip is measured in milliseconds.
04
UX research & design
Design Thinking driven: discovery interviews, prototype validation, Figma systems. Targeted at the user-facing half of every technical decision, with research methods that hand off cleanly to engineering.