At University of Missouri-Columbia, I built an MCP server that lets LLM agents call tools from the FuzzyLinguistics library. I also built an evaluation benchmark for an explainable OLAP navigation system; it reached closest to the target using one LLM call per task, compared with about 12 for LLM-only agents.
In my student technical assistant work at University of Missouri-Columbia (CAFNR), I processed drone imagery into elevation models used to measure crop height for the Corn Digital Twin. I’m also building a TabM model on PlanetScope imagery to predict where and when corn needs nitrogen.
At Indian Space Research Organization (Space Applications Centre), I validated Aerosol Optical Depth against other satellite and reanalysis data and automated multi-sensor processing pipelines. My projects include reinforcement learning for drone spectrum management, model quantization benchmarks, and a system for detecting unsupported claims in LLM answers.

