-
ML
Fine-tuned Transformer models and engineered a hybrid CNN + OLS pipeline for art market time-series forecasting; built a hedonic pricing index for investment valuation under volatile conditions.
-
Multimodal
Trained large-scale models on image + text description datasets for automated artwork classification and valuation; owned the full ML lifecycle from data labeling to model evaluation.
-
MLOps
Dockerized and deployed ML microservices; architected AWS Redshift pipelines and Apache Airflow DAGs to automate ETL from S3 for marketing attribution and financial reporting.
As a Data Scientist on Masterworks' Research team, I developed the proprietary ML valuation models and ETL pipelines that powered the platform's art market analysis — the core technology behind the company's growth to $882M AUM and 800,000+ members, as covered by ARTnews, CNBC Select, and Morning Brew.