Oct 2025 to present · Rochester, NY
RAG Engineer
Kate Gleason College of Engineering
Leading an AI platform that makes college data easier to reach. 92 percent answer relevance and 94 percent answer correctness.
The college has a lot of information spread across a lot of pages, and students end up in a ticket queue for answers that already exist somewhere. I lead the platform that fixes that.
I scrape college wide data through APIs and run it through a retrieval pipeline built with Python, Qdrant clusters, and Railway. Most of the real work is retrieval quality, chunking, and evaluation.
The current numbers are 92 percent answer relevance and 94 percent answer correctness on the questions advisors actually get. A Reeve Fund grant of two thousand dollars supports the deployment.