Webinar: AI for Drug Discovery

About this session

A webinar with Paul Hawkins, Principal Product Evangelist

Agentic AI is poised to reshape how molecules are designed, but an agent is only as good as the science beneath it. Train powerful AI tools on weak data using a crude model of molecular structure, and you don't get breakthroughs; you get wrong answers, faster.

This session explains how OpenEye, Cadence Molecular Sciences is developing solutions that pair agentic AI with principled simulation, curated data, and physics-based molecular representation to build workflows that do genuinely good science at scale, and free experts to focus where they matter most.

What you'll learn

• Why agentic AI delivers wrong answers faster when the underlying layers are weak
• How data quality makes or breaks a model
• The critical importance of prediction uncertainty in when working with agents
• How molecular representations from OpenEye, Cadence Molecular Sciences are the foundation of a wide range of AI models for molecular properties
• Why a SMILES string is not a molecule, and how shape and electrostatics can teach a model to "think like a receptor", resulting in models that increase the reliability and effectiveness of Human-In-The-Loop agentic workflows