Riya Gupta

I am a Senior ML Researcher working at the intersection of multimodal learning, medical imaging, and reliable AI systems. My current work focuses on medical imaging data at HPI, backed by Sinai, where I am building foundational models, vision transformers and LLM-based healthcare systems for clinical and scientific discovery.

Broadly, I am interested in building reliable AI systems for high-stakes domains with limited data availability. My work spans multimodal learning, representation learning, LLM evaluation, guardrails, agentic systems, medical imaging, and applied ML systems.

Previously, I worked on blood cancer research and developing search-based systems at HemeAI, backed by MSKCC; LLM-based guardrails, prompt-injection detection, evaluation, and safety infrastructure at Goldman Sachs - AI Research; and ASR and voice recognition systems at Silicon Labs, acquired by Texas Instruments. I completed my MS in Electrical & Computer Engineering at Columbia University, where my academic work spanned computer vision, NLP, speech, and reinforcement learning. I love to engage with the mathematics behind machine learning algorithms and models along with advances in robotics.

Outside research, I enjoy hiking and rock-climbing, both of which I have recently grown to love. I am a trained Hindustani classical singer and currently getting trained in Carnatic vocals and Bharatanatyam. I also spend time painting and exploring improv.