TerraForm stack for deploying ChatQnA using Vertex AI Vector Search

[arun-gupta]

ChatQnA is one of the GenAI examples. It is a chatbot for question and answer through retrieval augmented generation (RAG). All details about the sample are available at https://github.com/opea-project/GenAIExamples/tree/main/ChatQnA.

This example uses Postgres database with pgvector plug-in as a microservice. There is a need to show how Vertex AI Vector Search can be used instead of the built-in database. Any of the deployment mechanisms (Docker Compose, GKE, or Cloud Run) can be used.

This is a specific issue for Google Cloud for the mega issue at https://github.com/opea-project/GenAIExamples/issues/427.

[kevinintel]

Vertex AI Vector Search is on Google Cloud, but we don't have the cloud resource

[chickenrae]

@poussa Does you note solve this issue? Can it be closed?

[poussa]

Not done, can't be closed. Can be assigned to me if nobody else does not volunteer.

[chickenrae]

I've assigned to you @poussa , thank you for your help.

[xiguiw]

@poussa What's the status now?

[yinghu5]

Hi @poussa and All Just remind for any update, thank you!

[poussa]

I have no update. That is, I have not been able to work on it.

[yinghu5]

Hi @poussa Thank you a lot for the reply. As the issue open for a while, we close it for now. Please feel free to continue the work if needs. Some terraform deploy examples were listed in https://opea-project.github.io/latest/GenAIExamples/ChatQnA/benchmark/accuracy_faqgen/README.html