Skip to content

v2.9.0

Choose a tag to compare

View all tags
@placerda placerda released this 26 Jun 03:23
· 169 commits to develop since this release
v2.9.0
1d84fdb
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Adds Foundry IQ as a selectable retrieval backend for GPT-RAG.

This release lets operators run GPT-RAG retrieval through Foundry IQ by setting RETRIEVAL_BACKEND=foundry_iq. It also adds support for native user permissions through OBO headers and supports Pattern B authorization with filterAddOn for custom security fields.

What changed

  • Added Foundry IQ retrieval backend support.
  • Added OBO header support so Foundry IQ can evaluate native user permissions.
  • Added filterAddOn support for Pattern B deployments that use custom security fields.
  • Fixed the minimal-reasoning retrieve API path to use explicit semantic intents.

Why it matters

Operators can now choose between Azure AI Search and Foundry IQ retrieval patterns without changing the application contract. This enables deployments that need Foundry IQ Knowledge Bases, Knowledge Sources, and native permission evaluation.

Validation

Validated in a live Azure environment:

  • GPT-RAG provision and deploy completed successfully.
  • Foundry IQ direct retrieve returned HTTP 200 using generated Knowledge Base and Knowledge Source names.
  • Orchestrator smoke endpoint returned HTTP 200.
  • Orchestrator logs confirmed retrieval through Foundry IQ.
  • Frontend health, ingestion docs, and orchestrator docs endpoints returned HTTP 200.

References

Assets 2
Loading