datacraft workshop "Neuro-symbolic AI: The next step after LLMs, agents, and generative AI"

Neuro-symbolic AI: The next step after LLMs, agents, and generative AI By: Stéphanie BIRON (CEO of Auxo Dynamics, Data & AI Engineer specializing in semantics for complex systems)

This presentation will explore neuro-symbolic AI and its potential to transform how artificial intelligence is used in the corporate world.

The workshop will begin with an introduction to the core principles of this approach: how to bridge the learning power of neural networks—the backbone of modern machine learning and LLMs—with the rigor of symbolic systems, which are based on explicit rules, knowledge graphs, ontologies, and reasoning mechanisms. This first part will highlight how neuro-symbolic AI aims to overcome the current limitations of generative AI, particularly regarding explainability, reliability, knowledge structuring, and controlled reasoning.

The presentation will then demonstrate how this "marriage" between statistical learning and symbolic representation paves the way for AI systems that are more robust, interpretable, and better suited for complex business environments. Several concrete use cases will be showcased to illustrate the benefits of the neuro-symbolic approach in corporate settings: leveraging business expertise, decision support, reasoning over complex systems, enriching knowledge graphs, and improving the reliability of generative AI applications.