🎓 MS ADS @ UChicago · Applied AI Leader & ML Engineer · Classical ML · LLM/RAG · GenAI · MLOps · ex-TELUS strategy & tech
Movie recommendation system built on MovieLens, comparing Popularity, SVD (scikit-surprise), and LightFM across ranking metrics to evaluate whether Factorization Machines provide meaningful lift over simpler collaborative filtering baselines. Prepared for ADSP 31018 Machine Learning II (Autumn 2025).
Jupyter NotebookRAG system for Netflix recommendations using OpenAI embeddings, GPT-4o, and Milvus vector store. Includes LLMOps monitoring via Langfuse, Chainlit UI, FastAPI backend, Prefect DAGs, and GCP Cloud Run deployment.
PythonExploration of MLOps tooling built around a cancer mortality prediction model. Demonstrates how DVC, MLflow, Evidently, and Prefect fit together to version data, track experiments, detect drift, and orchestrate pipelines to mirror production ML workflows. Prepared for ADSP 32021 Machine Learning Operations (Winter 2026).
HTMLThis project is to design end-to-end MLOps pipeline predicting CrossFit athlete total lift using Airflow for orchestration, MLFlow for experiment tracking, Feast for feature management, and CodeCarbon for emissions monitoring. Prepared for ADSP 32021 Machine Learning Operations (Winter 2026).
Python