Here's a quick rundown of some projects from my career journey, covering both academia and industry.
I developed an anomaly detection tool utilizing Autoencoders (AE) to help engineering teams automatically flag abnormal datasets within their simulation/test decks. Leveraging deep learning in this fashion allowed us to catch setup errors early.
Architected a production-grade Agentic AI system on Databricks, integrating RAG with a custom math engine and Python UDFs to automate complex numerical analysis and high-fidelity technical data retrieval via LLMs
Developed a model that speeds up the way we simulate how different fuels burn, allowing for more efficient testing and design
Applied a MLP model to a CCHP system, enabling faster optimization and analysis.