Build a unified semantic layer across datasets with multi-dataset Topics in Amazon Quick

Amazon Quick has introduced multi-dataset Topics, a feature that enables the creation of a unified semantic layer across multiple datasets, allowing business users to ask questions in natural language and receive answers directly from their data. This feature supports up to 12 datasets per topic and defines relationships between them, which the AI engine uses to construct SQL joins and return unified answers. The multi-dataset Topics feature is available in public preview and expands support to various data sources, including Amazon Redshift, Amazon Athena, and Snowflake. This development has significant implications for engineers building AI systems, as it enables more efficient and effective data analysis and querying.

⚡ Key Takeaways

• Multi-dataset Topics in Amazon Quick support up to 12 datasets per topic.
• The feature defines relationships between datasets, which the AI engine uses to construct SQL joins.
• Multi-dataset Topics are supported in Amazon Quick Sight, with data sources including SPICE, Amazon Redshift, Amazon Athena, Amazon S3 Tables, Snowflake, and Databricks.
• The AI engine interprets user intent, identifies relevant columns, and returns a unified answer based on defined relationships.
• The same multi-dataset topic can be used for building analysis or answering questions using the chat agent.

The introduction of multi-dataset Topics in Amazon Quick has significant implications for engineers building AI systems, as it enables more efficient and effective data analysis and querying. This feature allows business users to ask questions in natural language and receive answers directly from their data, without requiring extensive technical expertise.

✅ Practical Steps

1. Create a multi-dataset Topic in Amazon Quick Sight by adding up to 12 datasets and defining relationships between them.
2. Use the chat agent to ask questions in natural language and receive unified answers based on the defined relationships.
3. Leverage the multi-dataset Topics feature to build analysis and answer questions using a single topic.

Want the full story? Read the original article.