A selection of data engineering and analysis projects I've worked on.
Architecting enterprise OCR pipeline designed to process millions of historical images annually. Multi-stage preprocessing with illumination correction and Google Document AI integration.
Redesigned genetic optimizer using Numba JIT and vectorized NumPy, reducing processing time from 1-2 minutes to 13-20 seconds (6x improvement). Containerized FastAPI backend with Google Sheets frontend.
Image alignment pipeline using OpenCV with Gradio interface for local testing and deployment.
Python/LangChain workflow for interactive survey data analysis and written response data extraction.