AI agents give confident wrong answers without enterprise context layer, Snowflake responds

AI Agents Struggle With Fragmented Business Context

Enterprise AI agents face a critical production failure that has nothing to do with the models themselves. As organizations move from single-layer retrieval-augmented generation to hybrid retrieval architectures, AI agents are producing different answers from identical data depending on which system asks the question2

2

. Revenue means one thing in a business intelligence dashboard, something different in a SQL table, and something else entirely in an agent instruction. VentureBeat's VB Pulse Q1 2026 data shows hybrid retrieval intent tripling from 10.3% in January to 33.3% in March, marking the fastest-growing strategic position in the dataset2

2

.

"There are a lot of tools out there that you can ask questions, you get a very confident answer, but whether it's correct or not is different," said Christian Kleinerman, EVP of Product at Snowflake2

2

. The problem stems from business logic distributed across SQL, BI dashboards and agent instructions, with no single system owning the definition. When multiple agents or tools query the same underlying data, they reason over different schemas and return different answers.

Snowflake Horizon Context Targets the Enterprise Context Layer

Snowflake unveiled Horizon Context at its annual Summit conference in San Francisco, a new set of semantic and metadata management capabilities currently in preview1

1

. The offering, launched as part of Horizon Catalog, collects metadata from across an enterprise's data estate, enriches it with business definitions, relationships, lineage and governance information, and makes that context available across AI and analytics systems1

1

.

Artin Avanes, head of core data platform at Snowflake, explained that these capabilities build on the company's acquisition last year of Select Star, a metadata management startup known for its integrations with database systems such as PostgreSQL and MySQL, business intelligence tools like Tableau and Power BI, and data pipeline tools such as dbt and Airflow1

1

.

Horizon Context pulls metadata from Postgres, SQL Server, Tableau and Power BI into the Horizon Catalog, ensuring every agent, BI tool and external system draws from the same governed definition rather than reasoning independently over a raw physical schema2

2

. Semantic View Autopilot automatically creates and refines semantic views over time, extending curated business logic without requiring ongoing manual effort.

Cortex Sense Adds Platform-Derived Intelligence

Alongside Horizon Context, Snowflake introduced Cortex Sense, a platform-derived layer that automatically builds and enriches context from customer data and usage patterns on an ongoing basis2

2

. Kleinerman described it as improving the default experience before any explicit curation has happened. "Think of Horizon Context as everything that is explicit and declared by customers, and Cortex Sense is anything that is implicit and derived by us," Kleinerman explained2

2

.

The two layers connect to Snowflake's existing retrieval infrastructure. Cortex Search, the company's RAG implementation, plugs into both CoCo and Cowork as a tool, so context enriched by either layer flows into retrieval workflows2

2

. While Horizon Context is a Snowflake technology, the company is tying it to the Open Semantic Interchange, making customer-declared definitions portable across third-party catalogs and tools.

Semantic Layer Governance Becomes Critical for Agentic AI

The semantic layer is becoming the foundation for trusted agentic AI because headless agents asking thousands of questions simultaneously have zero tolerance for inconsistent data definitions4

4

. Dave Mariani, chief technology officer of AtScale, emphasized that while you want large language models to be creative as research assistants, "you don't want them to be creative about the definition of revenue or customer"4

4

.

Carl Perry, head of analytics at Snowflake, noted that semantics can no longer live inside BI tools sitting above the data. "When you start using AI agents, which are not in BI tools and they intentionally are running autonomously and driving workflows, that model has to be where the agents are residing," Perry said4

4

. Organizations that built coherent-looking enterprise data models on top of fragmented underlying data got away with it when humans were in the loop. AI agents are far less forgiving.

A Crowded Field Targets the Same Problem

Snowflake joins an increasingly crowded field of vendors targeting the enterprise context layer. Microsoft has opened its Fabric IQ business ontology via MCP so any vendor's agent can draw from a shared governed definition2

2

. Redis launched Iris, a context and memory platform that sits between agents and their data, built on a storage engine redesigned for agent-scale retrieval volumes. Pinecone is repositioning from vector database to knowledge engine with Nexus, which compiles enterprise data into task-specific artifacts before agents ever query them.

Devin Pratt, research director at IDC, told VentureBeat that Snowflake is headed in the right direction. "Agents are only as good as the data and semantics behind them, so the context layer, not the model, is the thing to watch right now," Pratt said2

2

.

The Battle for the Intelligent Client and AI Back End

The larger strategic question extends beyond data silos to who owns the new intelligent client and the AI back end that makes it useful3

3

. The new client is the agent-based system of engagement—Snowflake CoWork and CoCo, Databricks Genie, Microsoft's Copilot, Google's Gemini Enterprise and others. These clients require what analysts call the system of intelligence: a back end that models enterprise data, business rules and tacit organizational knowledge in a way that both humans and agents can understand and act upon.

The key premise is that the system of intelligence ingests more than data and metadata from pipelines, catalogs and connectors. It also learns from the agentic client through skills, artifacts, semantic views, query history, agent actions and human reasoning traces3

3

. Those signals become feedback into the intelligence layer and help turn individual productivity into organizational intelligence.

For CIOs, Horizon Context should reduce operational complexity because it will provide a governed map of their organization's data estate, said Stephanie Walter, practice lead of the AI Stack at HyperFRAME Research. "The value is not simply cataloging where data lives; it is giving AI systems the metadata, lineage, permissions, and business context needed to retrieve the right data safely," Walter said1

1

.

Governance Becomes Accelerant Instead of Blocker

The fundamental shift is reframing data governance from a blocker to infrastructure that makes it safe to open data broadly4

4

. Mariani warned that enterprises tempted to lock everything down first and open doors later have the model backwards. Shadow AI proliferates precisely because organizations withhold data access from people who need it most. "You've got to say everything is open and we have to be confident that our governance infrastructure can make sure that the data doesn't get into the wrong hands," Mariani said4

4

.

Bob O'Donnell, founder and chief analyst of TECHnalysis Research, noted that many organizations recognize effective AI initiatives require a strong data foundation, but the complexity of consolidating and organizing enterprise data has made unified platforms and an enterprise context layer increasingly important advantages. Organizations are moving away from centralized data warehouse strategies and toward federated data approaches that allow data to remain in place while being connected and analyzed across systems. Traditional semantic layers were designed for structured data, but modern AI environments must also incorporate semi-structured and unstructured information through knowledge graphs that agents can query across tools like Slack and Jira.