Snowflake and Amazon QuickSight enhance AI-powered BI with semantic views
AWS ML Blog|Written by: 이현민|Jun 28, 2026|Updated Jun 29, 2026|1 views|
★★★★☆
Snowflake and Amazon QuickSight have introduced semantic views to address data inconsistencies and improve the reliability of AI-powered business intelligence. This new feature integrates business definitions directly into the data layer, ensuring uniform interpretation across all applications and reducing AI hallucinations.
Snowflake, in collaboration with Amazon QuickSight, has rolled out semantic views designed to bridge the "last-mile gap" in data analytics. This innovation aims to resolve common issues where different dashboards or AI agents present conflicting data, a problem often stemming from business logic being embedded within individual applications rather than at a shared data layer. By centralizing these definitions, the new feature seeks to enhance trust in analytics and accelerate decision-making processes for organizations.
A semantic view functions as a Snowflake schema object that directly attaches business definitions—including tables, relationships, metrics, and dimensions—to raw data. This means any downstream application querying the semantic view automatically inherits these standardized definitions, ensuring that both AI and traditional business intelligence systems interpret information consistently. This approach is critical for mitigating the risk of AI hallucinations, where AI models generate plausible but incorrect information due to ambiguous or inconsistent data inputs. The integration allows for a unified understanding of data across an enterprise's analytical ecosystem.
For users, developers, and enterprises, this development means a more reliable and governed data environment. Teams can now ask natural-language questions through tools like Cortex Analyst against a data layer that guarantees consistent business logic, leading to more trustworthy answers. Semantic views also come with object-level access controls, enabling authorized and governed usage across SQL, BI, and AI endpoints, similar to how tables and views are managed. This standardization and governance are expected to streamline data operations, improve the accuracy of AI-driven insights, and foster greater confidence in data-driven strategies.
— AIDEN Editorial Team · Reviewed by 이현민
What this means for the market
This development signifies a crucial step towards more reliable and trustworthy AI applications in the global market. By standardizing data definitions at the source, it addresses a fundamental challenge in enterprise AI adoption: data inconsistency leading to unreliable outputs. For developers, it simplifies the creation of AI and BI tools that can draw from a single, consistent source of truth, while for enterprises, it promises faster, more accurate decision-making and reduced operational friction in data management.
How this issue is unfolding
Companies are increasingly leveraging vast amounts of data to gain business insights and make AI-driven decisions. However, inconsistencies between data sources and fragmented business logic often undermine the reliability of AI systems, leading to issues like 'AI hallucinations.' Against this backdrop, AWS and Snowflake are introducing semantic views to integrate business definitions at the data layer, enabling all applications to interpret data consistently. This advancement aims to strengthen data governance and enhance the accuracy of AI-powered business intelligence systems.