Had a call with Jérémy Ravenel on future Ontologies and Knowledge Graph

Had an insightful conversation with Jérémy Ravenel about using ontologies as the foundation of next-generation AI systems. This was a continuation of my earlier talk with Voxel51 on similar themes.

We explored how ontology-driven AI can create more robust, interpretable, and context-aware systems. Key topics included:

• Enhancing knowledge representation in large language models — structuring domain knowledge so LLMs can reason over it more effectively
• Improving reasoning capabilities — moving beyond pattern matching toward structured inference
• Facilitating contextual information retrieval — using ontological relationships to return more accurate and relevant results
• Exploring entity relations — mapping how different concepts connect and influence each other across domains

A big part of the discussion centered on picking the right framework to build ontologies. As they say, “a good beginning is half done.” The choice of ontology framework shapes everything downstream — from how knowledge is represented to how agents can traverse and reason over it.

This ties directly into my broader interest in knowledge graphs and how they can power the next wave of AI infrastructure.

Read the original post on LinkedIn.