At Euphoria GenX, I engineered computer vision and NLP preprocessing pipelines that increased throughput by 35%. I also trained and fine-tuned deep learning models, improving F1 score by 8.5% over legacy baselines.
I built a RAG knowledge engine that searches 500+ research papers with sub-450ms retrieval latency and 93.4% factual answer precision. My projects also include multi-camera traffic analytics and satellite imagery analysis, using tools such as PyTorch, FastAPI, and TensorRT.

