2026 workshop program
Democratizing Astrophysical Research with AI Agents: Lessons from Heliophysics
The work
Abstract
Astrophysical research is often bottlenecked by friction in data access, the runtime cost of complex numerical schemes, and the heterogeneity of data formats — overhead that consumes a substantial fraction of researchers’ time and effort. Recent advances in AI agents, together with emerging standards such as the Model Context Protocol (MCP) and agent skill systems, offer an unprecedented opportunity to reduce this friction by exposing data, analysis, and simulation workflows through standardized agentic pipelines. In this talk, I will present several such pipelines that have substantially accelerated research in heliophysics, and describe a general methodology for extending this approach to other subfields of astronomy.