Session 11-C
From Noise to Value: Unlocking Domain-Specific AI with Unstructured Data
In highly regulated industries like insurance and finance, critical business knowledge is often buried in unstructured formats, from claims notes to compliance documents.
In this session, Michael Mocanu (Liberty Mutual) and Or Zabludowski (Flexor) will explore how leading data teams transform messy, unstructured data into trusted, AI-ready assets, and why that’s essential for building scalable, domain-specific AI systems.
Key discussion points will include:
- The Evolution of the Feature Factory: How GenAI is reshaping feature engineering workflows, and why balancing flexibility with focus is critical for data science teams.
- Why General Co-Pilots Fall Short: The limits of generic GenAI tools in complex environments, and the growing need for domain-specific AI agents that understand context and ensure compliance.
- Structuring for Success: Why the formatting and standardization of unstructured inputs are foundational to enabling usable, trusted, and scalable AI.
Join us to learn how leading data teams are transforming noisy, unstructured data into strategic assets, and driving AI initiatives that deliver real, governed impact.
Speaker
MICHAEL MOCANU
Senior Director, Liberty Mutual
OR ZABLUDOWSKI
Co-Founder & CEO, Flexor