aiconfig-segments

Segments group contexts like users and organizations by shared attributes. Create a segment with a unique key, name, and description, then attach rules using LaunchDarkly operators to define inclusion criteria—for example, users with beta_opted_in set to true—so you can reuse consistent targeting across AI Configs.

Yes. Segments enable rule-based targeting that works across environments. Define multi-criteria rules once in a segment, then reference that segment in AI Config targeting rules to apply consistent logic without duplicating configuration across environments.

You need a LaunchDarkly API token with segments:write permission and Python-based operations to construct segment payloads and patch segment rules. This allows you to programmatically create, update, list, view, and delete segments.

Segments let you define inclusion criteria once as a reusable logical unit, then reference that segment across multiple AI Configs. This eliminates repeating the same targeting rules and keeps definitions consistent when requirements change.

Segment rules support LaunchDarkly operators to match context attributes. You can build multi-criteria definitions—such as matching users where beta_opted_in equals true or organization tier matches a specific value—to control which contexts the segment includes.

Yes. The segment management features let you list all segments, view details of specific segments, and query segment information across different environments, making it easy to audit and reference segments in your AI Config rules.