jonsnow User

I am a Bachelor of Technology candidate in Electronics and Communication Engineering at IIIT Allahabad with a minor in Economics and Finance, focused on machine learning, audio processing, and fraud-risk modeling. I have built end-to-end research and production-style pipelines—spanning dataset creation, feature engineering, model training, and evaluation—for projects that include audio-enhanced handwritten math transcription, transformer-based audio classification, and hybrid fraud risk scoring.

My work demonstrates strong practical results: high-accuracy environmental sound recognition, robust fraud-detection lifts, and novel dual-encoder Transformer approaches for InkML→LaTeX transcription. I contribute in leadership roles to campus events and bring hands-on experience with Python, TensorFlow/Keras, PyTorch, Hugging Face, scikit-learn, and cloud and collaboration tools, seeking opportunities to apply ML and audio expertise to impactful problems.