Biography
Sarthak is a Senior R&D Staff at the U.S. Department of Energy's Pacific Northwest National Laboratory, where he builds agentic AI systems for energy infrastructure planning, grid resilience, and federal permitting. His work helps public agencies and energy stakeholders respond to AI-era load growth, data center demand, and the need to plan, permit, and deliver new infrastructure faster.
At PNNL, he has contributed to DOE initiatives on grid modernization, including the Resource Adequacy Report on U.S. grid reliability, security, and capacity. He also developed GridCoPilot, one of the first LLM-based frameworks for long-term grid planning and reliability analysis.
He is a key contributor to PermitAI, a generative AI initiative for accelerating federal permitting and NEPA reviews for energy infrastructure. Related work includes the NEPA-MCP Toolkit, a Model Context Protocol toolkit that gives AI agents real-time, source-cited access to federal permitting data sources and regulatory context. It also includes DraftNEPABench, a PNNL-OpenAI collaboration benchmarking coding agents for drafting environmental review documents.
He is also a core member of the IEEE Power & Energy Society Task Force on Foundation Models, where he contributes to advancing thought leadership on the application of foundation models in power system analysis, operations, and planning.
Prior to joining PNNL, he conducted research at Georgia Tech on optimization of electric vehicle charging infrastructure and pricing markets. His research on charging infrastructure and pricing mechanisms has been featured by AAAI, Harvard Business School, Microsoft, and The New York Times.