We’re excited to announce that ServiceNow is joining the Open Semantic Interchange (OSI). It's an industrywide initiative where Snowflake, Databricks, and many other data leaders and pioneers are standardizing how semantic metadata is defined and integrated across systems.
Let's explore why this is such a big deal.
Every conversation about enterprise AI eventually arrives at the same question: How does the model know what your business jargon means?
A customer relationship management (CRM) customer isn’t necessarily the same as a customer in billing. Active revenue means one thing to finance but something different to sales. Multiply those mismatches across hundreds of systems, and you have the reason most AI pilots stall before production.
The answer isn’t another data warehouse or another model. It’s semantics—definitions, metrics, ontologies, and relationships—that travel with the data and tell AI what it’s looking at. To be most effective, metadata must be accessible, governed, and portable, not an afterthought or a black box.
The stakes rise sharply the moment AI acts rather than merely describes or recommends. The wrong recommendation can be ignored. The wrong action has consequences.
Every large enterprise has multiple data catalogs, semantic layers, and sources of truth. This isn’t by design. It happened organically as new tools and platforms were added to the stack and each one brought its own operational model. Each of these semantic layers reflects intentional work. They just don’t connect.
Organizations aren’t going to move to a single tool any time soon. But what they can do is connect semantics, govern them, and make them interoperable. This allows AI to find trusted information and understand the metrics, policies, and definitions that are best suited for the question or action at hand.
ServiceNow is where many of our customers build their AI governance strategies. It’s where enterprise context is curated, governed, and turned into action.
ServiceNow Context Engine, ServiceNow Data Catalog (acquired from data.world), and Pyramid Analytics (also recently acquired) make this possible. They form an enterprise ontology and semantic layer that drive AI and the business forward across data platforms, business intelligence tools, applications, and AI.
That’s what an open standard like OSI is designed to achieve.That’s what an open standard like OSI is designed to make possible.
We’re joining OSI’s working groups such as Ontology Representation, Catalog Integration, Metrics and Expression Language, and Composability. The same thesis runs through each: Semantics are most valuable when they’re connected to action.
Much of the interchange’s work is focused on how semantics flow into analytics. ServiceNow also brings its perspective of how those same semantics flow into workflows, AI agents, and operational systems, where business outcomes happen.
If you’re investing in a practice around semantics and context, ServiceNow joining OSI is good news. The work you put into defining your business doesn’t have to be redone every time a new tool enters your stack.
Our semantic layer and Context Engine will optionally integrate with existing ones and emerging standards. Your definitions can travel, your governance can hold, and your AI can act on context it understands.
Semantics are how AI learns the shape of your business. ServiceNow is glad to be part of the group working to make it more portable and powerful, and to put those learnings into autonomous action.
Find out how ServiceNow can help clarify your data strategy.
