Uploaded April 2026 | Updated September 2026, 1 week ago
From SAS to Open Source: Building an AI-Ready Clinical ADaM Pipeline
In this webinar, Appsilon discusses what it takes to move clinical ADaM programming from traditional SAS-based workflows toward a more scalable, open-source approach.
The session covers the real bottlenecks in preparing ADaM deliverables, why code generation alone is not enough, and how a metadata-driven, component-based framework can help reduce validation effort, improve reuse, and create a stronger foundation for AI-supported clinical programming.
You’ll also see a live walkthrough of mightyverse, an open-source framework for ADaM automation, including how it uses specifications, reusable components, topology generation, and submission-ready outputs.
What you’ll learn:
- Why traditional ADaM programming can become difficult to scale
- How metadata-driven workflows help create a single source of truth
- How mightyverse supports reusable, validated components
- Where AI can realistically fit into clinical programming
- Lessons learned around governance, validation, adoption, and stakeholder alignment
Speakers:
Matthew Phelps, Clinical Data Scientist, Novo Nordisk
Ryszard Szymański, Staff Engineer, Appsilon
Moderated by Vedha Viyash, Appsilon
Learn more about Appsilon: appsilon.com
From SAS to Open Source: Building an AI-Ready Clinical ADaM Pipeline
In this webinar, Appsilon discusses what it takes to move clinical ADaM programming from traditional SAS-based workflows toward a more scalable, open-source approach.
The session covers the real bottlenecks in preparing ADaM deliverables, why code generation alone is not enough, and how a metadata-driven, component-based framework can help reduce validation effort, improve reuse, and create a stronger foundation for AI-supported clinical programming.
You’ll also see a live walkthrough of mightyverse, an open-source framework for ADaM automation, including how it uses specifications, reusable components, topology generation, and submission-ready outputs.
What you’ll learn:
- Why traditional ADaM programming can become difficult to scale
- How metadata-driven workflows help create a single source of truth
- How mightyverse supports reusable, validated components
- Where AI can realistically fit into clinical programming
- Lessons learned around governance, validation, adoption, and stakeholder alignment
Speakers:
Matthew Phelps, Clinical Data Scientist, Novo Nordisk
Ryszard Szymański, Staff Engineer, Appsilon
Moderated by Vedha Viyash, Appsilon
Learn more about Appsilon: appsilon.com
