Create SQL Segments Using Natural Language

In this release, Fusion Unity has introduced a Natural Language to SQL capability powered by an AI agent, enabling users to create SQL-based segments by simply describing their criteria in plain English. The agent translates user input into SQL queries and supports iterative refinement through a conversational interface. This significantly accelerates segment creation while maintaining the flexibility and power of SQL-based segmentation within Fusion Unity’s advanced audience framework.

Business Benefits:

  • Speeds up SQL segment creation by eliminating the need to write queries from scratch
  • Lowers the barrier for users with limited SQL expertise while still supporting advanced use cases
  • Improves productivity by enabling rapid iteration through conversational query refinement
  • Reduces time to insight and activation for complex segmentation scenarios

Steps to enable and configure

You don't need to do anything to enable this feature.

Tips and considerations

How to Start Using

  1. Click on the "Ask Oracle" button on the segment list page to start interacting with the agent.
  2. Describe your desired audience in plain English, for instance "Contacts with a job level of 'Executive' who are in the Sales department and who work for an account that has a subscription that started in the past six months."
  3. Review the resulting SQL and iterate as needed. 
  4. When you are satisfied with the result, instruct the agent to create your segment.

Tips and Considerations

  • Best suited for technical users or marketers needing advanced segmentation beyond the visual builder
  • Review and validate generated SQL to ensure accuracy and performance
  • Use iterative prompts to refine logic rather than rewriting queries manually from scratch
  • Over time, rely more on the agent to accelerate increasingly complex segmentation workflows
  • Ideal for jump-starting query creation, even if final tuning is done manually