Generative BI in Amazon QuickSight

Let me clarify your questions about Generative BI in Amazon QuickSight:

1. Q topics in QuickSight do not inherently use generative AI or LLMs. The classic QuickSight Q&A experience, which uses the Q search bar, is separate from the newer Generative BI capabilities. However, Amazon QuickSight now offers a Generative BI Q&A experience powered by Amazon Bedrock large language models (LLMs), which is distinct from the traditional Q topics.

2. When users ask a question using the classic Q topic feature in QuickSight, the system does not use generative AI to create SQL queries. Instead, it relies on natural language understanding to interpret the question and match it to pre-configured datasets and fields. The system then generates a visual answer based on the interpreted query, using the appropriate fields, aggregations, and datasets to create the visualization.

The newer Generative BI capabilities in QuickSight, powered by Amazon Bedrock LLMs, offer more advanced features. These include the ability to build visuals, create complex calculations, and refine visualizations using natural language prompts. This generative AI-powered system can interpret more complex queries and generate appropriate visualizations or calculations without relying on pre-configured mappings.

It's important to note that while both the classic Q topics and the new Generative BI features aim to provide natural language interfaces for data analysis, they operate differently. The Generative BI capabilities offer more flexibility and power in interpreting and responding to user queries, leveraging the capabilities of large language models to understand context and generate appropriate responses.

Sources
Asking questions with Amazon QuickSight Q - Amazon QuickSight
Generative BI dashboard authoring capabilities now available in preview for Amazon QuickSight Q customers | AWS Business Intelligence Blog
Generative BI dashboard authoring capabilities for Amazon QuickSight (Preview)