Bio

Harshit Singh is an AI Researcher and Engineer working across generative AI, agentic systems, and environmental modeling. He is currently working on the Natural Capital Project at Stanford, where he develops LLM-driven workflows for the InVEST ecosystem to enhance automation, data integration, and sustainable development research. He is also building HarshanAI, an emotionally intelligent voice-AI companion.

Previously, he worked at Amazon Web Services, contributing to Bedrock Flows and AgentCore for large-scale generative AI systems and at the MIT-IBM Watson AI Lab, leading DiffuseKronA as first author and advancing parameter-efficient methods for personalized diffusion models. He has also supported climate and energy research at the Center for Global Sustainability, University of Maryland through the development of G-MAST, a global methane abatement solutions database. His work emphasizes practical innovation, scalable AI systems, and applying machine learning to real-world societal and sustainability challenges.

Contact

  • Work
    harshan@stanford.edu
    University - Staff Department: Woods Research Natural Capital Project Position: Research Assistant

Additional Info

  • Mail Code: 5020