Rethinking the Semantic Layer- Part II: The Builders Response

​Semantic layers were built for a world where humans asked business questions, and the semantic layer helped translate those questions into a SQL query over a warehouse. But now that more business context is moving directly into code, agents can read schemas, inspect code, generate queries, follow lineage, and reason across systems.

In our last webinar we argued that most of what a semantic layer provides for agents already lives in well-written pipeline code, while acknowledging the places where the equivalence breaks. This fireside chat is the next move in that conversation: we want to hear from people building semantic layer products about what they are actually seeing.

So what is left for the semantic layer to do?

To discuss the future of semantic and context layers, we will be joined by:

• ​Moderator: Jacopo Tagliabue (CTO, Bauplan)

• ​Matthew Mullins (CTO, Coginiti)

• ​Simba Khadder (Lead Context Engine, Redis)

• ​Paul Blankley (Founder, Zenlytic)

​Together, we will push towards an everything-is-code data workflow, we will ask people who have spent years building in this space what they are actually seeing with customers, teams, and production systems.

The conversation will cover what each of these builders is seeing in customer deployments, where the semantic layer is being absorbed into code, where it is holding its ground, and what new primitives the agent-native stack still needs. The discussion will also leave room for audience questions on what is real, what is hype, what breaks in enterprise deployments, and what new primitives are still missing.

We will close with the question we keep coming back to: is the semantic layer's future a novel abstraction nobody has built yet, or a quiet reinvention inside other pieces of the stack?