Now
Last updated .
Senior Machine Learning Engineer at CVS Health.
Taking AI Evals for Engineers and PMs (Parlance Labs, Oct 10 – Nov 21, 2026).
Side projects
These are personal projects and bootcamp exercises. None of them are production systems.
- AI retail app and 372-persona shopping simulation (personal). A retail app plus a LangGraph simulation of 372 shopping personas. It generated 2.1M rows and 3+ years of simulated shopping data in 2 weeks.
- Retail ML platform (personal). A retail machine-learning platform using Supabase, dbt, Feast, DVC, MLflow, and Terraform.
- LLM post-training: Qwen2.5-1.5B LoRA SFT (personal). LoRA supervised fine-tune of Qwen2.5-1.5B on synthetic retail customer-service data (496 train / 30 held-out). Three adapter iterations beat the prompted base model on a custom rubric. Only stage 1 of 5 of the planned pipeline is done.
- Support-agent evaluation pipeline (bootcamp). Evaluation pipeline for a support agent, built with the OpenAI Agents SDK, an LLM-as-judge, Langfuse and OpenTelemetry tracing, and CI regression tests. The repository is private until after November 21, 2026.