Hey, I'm Jesal π β Senior QA Engineer who stress-tests AI systems for a living.
My stack spans PyTorch, TensorFlow, ONNXRT, OpenCV, LangChain β basically anything that runs a model, I've probably tried to break it. From 70B LLMs to real-time CV pipelines, if it ships, it passes my gate first. π
Hardware roots, software brain β started with PMSM motor controllers and EV systems, ended up profiling transformers. The throughline? Make it reliable or don't ship it. β‘ Currently overfitting to test automation. Occasionally gardening to reset the weights. π±