##  [Inside Via: Agentic AI Grounded in Regulatory Logic](/thought-leadership-articles/inside-via-agentic-ai-grounded-in-regulatory-logic) 

Most AI tools built for the regulatory compliance can find information but cannot be held accountable for it. At the core of [Via, our variation identification agent](https://www.freyrsolutions.com/meet-via-variation-identification-agent?utm_campaign=44276337-Via%20Launch&utm_source=Freyr%20Solutions%20Blogs), is a governed, ontology-driven knowledge graph that stores regulatory logic as a web of relationships, market-specific rules, and precedents, every one of them expert-seeded and version-controlled.

[**Watch Via in Action**](https://www.freyrsolutions.com/meet-via-variation-identification-agent?utm_campaign=44276337-Via%20Launch&utm_source=Freyr%20Solutions%20Blogs#video)

As [Praveen Bezawada](https://www.linkedin.com/in/praveen-bezawada/), CTO, Freyr Solutions puts it: ***"Where an LLM infers, the KM (knowledge manager) asserts."*** [Click here to read the article.](https://www.linkedin.com/pulse/from-knowledge-action-how-via-thinks-through-praveen-bezawada-lpxif/)