Semantic Views FAQs
What are Semantic Views?
Semantic Views provide a semantic layer built on top of ATTOM’s property data delivered through Snowflake. They provide AI and analytics tools with context about how ATTOM’s data should be understood and queried, including relationships, business definitions, metrics, synonyms and query logic.
Unlike a static semantic model, Semantic Views are actively evaluated and refined by ATTOM as usage and data needs evolve.
Why are Semantic Views important for AI?
Access to data alone does not mean an AI model understands how to use it correctly. Without the right context, AI can choose incorrect relationships, misinterpret fields or generate SQL that appears correct but produces an inaccurate result.
Semantic Views provide governed definitions and relationships that help AI systems interpret and query ATTOM’s property data more reliably.
Why use Semantic Views built by ATTOM instead of building my own semantic layer?
Building a semantic layer requires understanding ATTOM’s schemas, relationships, terminology, coded values, metrics and query logic.
ATTOM has already done that work. Semantic Views deliver ATTOM’s understanding of its property data alongside the data itself, reducing the need for customers to independently build and maintain that context.
How does ATTOM continually improve Semantic Views?
Semantic Views are actively evaluated and refined as data needs, business questions and use cases evolve. ATTOM continually reviews the semantic context behind its property data, including definitions, relationships, metrics and verified queries, to identify opportunities for improvement.
Want to Go Deeper on Semantic Views?
Explore how Semantic Views bring business context, governed logic and natural-language querying to ATTOM’s property data in Snowflake.
Can I ask questions about ATTOM’s property data in natural language?
Yes. When Semantic Views are connected to Snowflake Cortex Analyst and Cortex Agents, users can ask property data questions in plain English without needing to know the underlying table names, joins or SQL syntax.
Cortex Analyst uses the definitions and relationships within the Semantic View to translate the question into SQL and return a natural-language answer.
What information is included in Semantic View?
Depending on the underlying ATTOM property data product, Semantic Views can include table and field descriptions, synonyms, relationships, dimensions, sample values, governed metrics, verified queries and custom instructions that help AI understand how the data should be interpreted.
What ATTOM property data is available through Semantic Views?
Semantic Views are available across a growing range of ATTOM’s property data products, with coverage continuing to expand.
Can multiple Semantic Views work together?
Yes. A Snowflake Cortex Agent can connect to multiple Semantic Views. Cortex Analyst can evaluate an incoming question and determine which Semantic View is best suited to answer it, allowing users to interact with multiple ATTOM property data products through a common agent experience.
Do Semantic Views replace ATTOM’s underlying property data?
No. Semantic Views do not change or replace ATTOM’s underlying property data. They add context about what the data means, how it relates and how it should be queried, making it easier for people, applications and AI systems to get useful answers from the data.
How are Semantic Views priced?
Semantic Views are available as a licensed solution for supported ATTOM property data products. Once deployed in a customer’s Snowflake environment, usage is subject to the customer’s applicable Snowflake compute costs.