Catalyst Copilot is an AI-powered agent embedded in Amgen’s R&D knowledge management platform, Catalyst. It was built using Microsoft Copilot Studio, a low-code environment for creating intelligent agents and agentic workflows.

Amgen’s core challenge was that critical R&D knowledge—such as trial designs, molecule data, reports, PowerPoints, PDFs, and SharePoint content—was scattered across many digital locations. Drug developers often had to spend hours or even days hunting for the right information and the right people, which slowed down knowledge transfer and, ultimately, drug development.

Catalyst Copilot addresses this by:

• Ingesting and reasoning over Amgen’s curated R&D resources.
• Providing a Q&A style chat interface so users can ask questions in natural language, like they would with a colleague.
• Surfacing relevant information and experts faster, helping teams move from “Where is the information I need?” to “What do we already know?”

The broader goal is to shorten the cycle of discovery and support Amgen’s mission of getting medicines to patients more quickly by treating drug development as an information capability.

Amgen built and deployed Catalyst Copilot in about six weeks.

Several factors made this timeline achievable:

• Existing curated knowledge base: Amgen had already done the “hard work” of organizing and curating R&D content—lessons learned, articles, presentations, and other resources. This meant the AI could be applied to a well-structured information set.
• Use of Microsoft Copilot Studio: Copilot Studio is a low-code platform that lets teams build intelligent agents and custom workflows without starting from scratch. Amgen could rapidly configure, test, and refine the agent.
• Early adoption of Microsoft 365 Copilot: As an early adopter, Amgen already had experience with Microsoft’s AI ecosystem, which reduced friction in experimenting and deploying new solutions.
• Collaboration between IT and R&D knowledge teams: The technology team partnered closely with the R&D knowledge management group to align the agent’s capabilities with real user needs.

This combination of a trusted technology collaborator (Microsoft), a curated data foundation, and a low-code platform allowed Amgen to move from concept to a working agent in a relatively short period.

Catalyst Copilot is helping Amgen reimagine how R&D teams access and use knowledge across the organization.

Key changes in day-to-day work include:

• Faster access to knowledge: Instead of sifting through intranet links, folders, and documents, drug developers can ask questions in natural language and get synthesized answers. This reduces the “white space” time spent searching for information and people.
• Better use of collective intelligence: Catalyst Copilot reasons and infers across multiple fact sets, helping teams reach faster, more accurate conclusions and tap into the organization’s collective experience.
• Support for better science at scale: One of Amgen’s key AI metrics is “enabling better science at scale.” Catalyst Copilot contributes by helping automate parts of the scientific information process and distilling data into insights that researchers can act on.
• Improved content curation: The Q&A chat interface not only serves users but also gives the knowledge management team visibility into user intent, so they can refine and more precisely curate content over time.

Looking ahead, Amgen plans to extend this approach, using Copilot Studio as a common platform to deploy additional agents and custom workflows in other parts of the business, with a vision of a single user experience that routes people to the right agent for their needs.