AI is getting better at generating answers, but how do we know when those answers should be trusted?

Most AI systems reduce confidence to a simple percentage. Important information such as uncertainty, provenance, conflicting evidence, temporal validity, and compliance requirements is often lost along the way.

Join us for a hands-on workshop with Muntaser Syed, creator of the open-source jsonld-ex framework, as we explore new approaches for making AI honesty machine-readable.

During this session you'll learn how to:

✅ Capture confidence, uncertainty, and provenance alongside AI-generated facts

✅ Combine information from multiple AI systems using principled mathematical approaches

✅ Detect conflicts between competing model outputs

✅ Model regulatory compliance and consent lifecycles

✅ Build confidence-aware AI pipelines

Participants will work through multiple Google Colab notebooks and see practical examples of:

• Multi-source AI confidence fusion

• Trust discounting and uncertainty modeling

• Compliance-aware AI systems

• Knowledge extraction pipelines

• W3C JSON-LD interoperability

Whether you're building AI applications, data platforms, intelligent agents, or enterprise systems, this workshop offers a unique perspective on creating more transparent, trustworthy, and auditable AI solutions.

No GPU or semantic web experience required.

Prerequisites

• Basic Python knowledge

• Laptop

• Google Account

• Curiosity about trustworthy AI

About the Speaker

Muntaser Syed

Muntaser is a software engineer and creator of the open-source jsonld-ex project.

His work focuses on extending W3C JSON-LD standards with machine-readable confidence, uncertainty, provenance, and compliance metadata to help build more trustworthy AI systems.

Project links:

GitHub:
https://github.com/jemsbhai/jsonld-ex

PyPI:
https://pypi.org/project/jsonld-ex/

NPM:
https://www.npmjs.com/package/@jsonld-ex/core

LinkedIn:
https://www.linkedin.com/in/muntasersyed/