Customer facing documentation for Multi-region AI Gateway
https://gitlab.com/gitlab-org/gitlab/-/issues/462358
Duplicatescontinue work on that issue
Problem
Customers are very excited about our Multi-region AI Gateway support, however they do not understand how it works and exactly what benefits it provides (latency reduction? data residency? etc)
Requirements
-
Document Multi-region gateway support does NOT garuntee data residency, and explain why. -
Document how we decide which region queries go to. What logic determines which regional gateway a query from X location routes to -
We've got duplicated information on all the below pages which can make this confusing for GitLab staff as well as customers to understand, we should consolidate these as much as possible.
Possible doc locations
- Blue Print: Gateway https://docs.gitlab.com/ee/architecture/blueprints/ai_gateway/
- Dev Docs: AI Architecture https://docs.gitlab.com/ee/development/ai_architecture.html
- Blue Print: Cloud Connector https://docs.gitlab.com/ee/architecture/blueprints/cloud_connector/
- Duo Data usage https://docs.gitlab.com/ee/user/ai_data_usage.html
Timeline
We need answers to these questions quickly to support high value customer sales. It's ok to write it up here and @sselhorn can help us get it publicly documented.
Edited by Taylor McCaslin