AI/ML STIG Lecture Series
Artificial Intelligence & Machine Learning Science and Technology Interest Group
Location
Virtual
dates
9 November 2026
4:00pm ET
Community
AI/ML STIG
Type
Lecture
AI Tool Use for Astronomy, and MCP for Data Archives
Speaker
Dan Gause, Adele Plunkett (NRAO)
Abstract
Large Language Models (LLMs) offer a powerful mechanism for interacting with massive, growing astronomical data archives using natural language. There are, however, several challenges to their scientific use for data-driven astronomy, rooted in the distinct schemas, query services and conventions of each archive. In this talk we present STABLE (“Summer Team for Astronomical Benchmarking & LLM Engineering”), an initiative supported by a 2026 CosmicAI Seed Funding Award, which pursued three complementary aspects of AI tools: rigorous benchmarking, infrastructure enhancement, and AI agent development. The primary infrastructure deliverable was MANNA (“MCP Architecture for NOIRLab, NRAO, and Additional Archives”), a Model Context Protocol (MCP) server that gives AI clients standards-based access to astronomy archives. As a component of this talk, we cover the conceptual and technical aspects of MCP components followed by a demo on building an MCP server. We also discuss examples of MCP servers in astronomy, detailing the development and evaluation of MANNA.
Meeting Connection
News Straight to Your Inbox
Subscribe to your community email news list
We will never share your email address.



