At Mixle, I collaborate across product and engineering teams on Event Mode, including event-scoped intent matching, recommendations, and mutual-match flows. I built and maintain PostgreSQL-backed REST APIs and shipped LLM-based enrichment grounded in live app and user data.
For independent clients, I architected an AI document intelligence platform for a legal client, combining OCR, structured extraction, and LLM summarization in a HIPAA-compliant design that cut attorney document-review time by 50%. I also designed a RAG pipeline with OpenAI GPT-4 and Chroma, and built backend services with Python and AWS.
At Moultrie (EBSCO Industries), I built and shipped deer re-identification using YOLO-based detection and embedding matching, and iterated on training based on evaluation findings. My work also includes software engineering model evaluation at Handshake AI Fellowship and data science research at Glycans for Regeneration Lab, UAB.
