I lead AI and data engineering at Vinotheque, building the data pipeline and analytics the business runs on.
- 🎬 AACC: an agent harness that turns your data into video presentations automatically (aacc-cli)
- 🍷 The data and analytics stack at Vinotheque
- 🧬 Contributing to TPOT, a 10k★ Python AutoML library that optimizes ML pipelines with genetic programming
- 🤖 Co-authored early work evaluating LLM code generation against genetic programming on program synthesis (LLMvsGP_comparison, GECCO 2025)
- 🛠️ Open-source ML utilities, like a memory-efficient Variational Autoencoder Imputer for scikit-learn
Before Vinotheque I did ML research at Cedars-Sinai, on AutoML and biomedical data.
- Comparing PushGP and GPT-4o on Program Synthesis with only Input-Output Examples, GECCO 2025
- GP and LLMs for Program Synthesis: No Clear Winners, GPTP XXII (arXiv:2508.03966)
- Pharmacokinetic and pharmacodynamic assessment of oral nicotinamide in the NEAT trial for early Alzheimer's disease, Alzheimer's Research & Therapy, 2025
📖 Full list on Google Scholar
Python · PyTorch · scikit-learn · SQL · TypeScript · R



