AI Indexing Cost and Storage Overview
To enable AI search, Codebeamer processes the project content by generating vector embeddings for the configured data.
When you enable indexing for a project:
• Codebeamer divides the configured content into smaller segments.
• Each segment is converted into a vector embedding, which allows the AI search to understand and retrieve relevant results.
 |
If identical content exists across multiple streams or working sets, Codebeamer creates only one embedding for that content to optimize resource usage.
|
AI Credit Consumption for AI Indexing
Indexing and maintaining searchable content consumes AI credits:
• Initial indexing: Creating embeddings for project content consumes AI credits.
• Ongoing updates: When project data changes, for example, when you create or modify items, new or updated content is processed again. This additional processing also consumes AI credits.
Storage and Billing
• All embeddings and their associated index structures are stored in a vector database used by the indexing service.
• This database uses disk space, which is billed in the form of AI credits.
Content Lifecycle and Storage Optimization
To manage storage usage efficiently:
• Codebeamer removes outdated content, such as previous versions of items, when it is replaced by newer versions.
• This cleanup applies across all streams and helps reduce the overall storage footprint of the vector database.
Key Considerations for Administrators
• Plan indexing scope carefully to control AI credit usage.
• Monitor content updates, as frequent changes increase processing costs.
• Be aware that storage costs are ongoing and depend on the volume of indexed content.