Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Deploying Two Different Chatbots While Sharing Infrastructure #2229

Open
CarlosPSR opened this issue Dec 12, 2024 · 2 comments
Open

Deploying Two Different Chatbots While Sharing Infrastructure #2229

CarlosPSR opened this issue Dec 12, 2024 · 2 comments

Comments

@CarlosPSR
Copy link

Please provide us with the following information:

This issue is for a: (mark with an x)

- [ ] bug report -> please search issues before submitting
- [ ] feature request
- [x] documentation issue or request
- [ ] regression (a behavior that used to work and stopped in a new release)

Minimal steps to reproduce

Any log messages given by the failure

Expected/desired behavior

Being able to deploy two differently tuned RAGs with separated data sources but sharing resources when possible.

OS and Version?

Linux (Ubuntu).

azd version?

azd version 1.9.0

Versions

Mention any other details that might be useful


Hi there,

First of all, I am very new to this field, so my apologies if I ask a silly question. I have successfully tuned and deployed the demo, and I am using it in production for a webpage. I am using the chatreadretrieveread method, with a specific system prompt, a few-shot examples, and targeting certain PDF documents.

What I need is to create another bot with a different system prompt and to use RAG with a completely different set of PDF documents (unable to access the first set). However, I would like to share services/resources with the first deployment as much as possible to reduce costs. I understand that this is possible, although I might need to create another App Service and at least a new index in the vector database. Please correct me if I’m wrong.

I’d really appreciate it if you could guide me or give me some hints on the feasibility and the best approach to achieve this, either by modifying the code or manually deploying services in the Azure Portal.

Thank you very much in advance,
Carlos

@pamelafox
Copy link
Collaborator

The general approach would be:

  1. Create a new azd environment with azd env new command
  2. Run the commands in https://github.com/Azure-Samples/azure-search-openai-demo/blob/main/docs/deploy_existing.md to tell that environment to reuse all the desired existing resources. Let me know if we're missing anything from there, it's very possible.

I've done #1 and #2 before, but not recently, and success may depend on exactly what you're going to share.

@CarlosPSR
Copy link
Author

CarlosPSR commented Jan 9, 2025

Hi @pamelafox ,

Thanks a lot for your answer. I followed those steps and encountered some trouble related to roles.sh script, as it returned errors trying to assign already existing roles. I guess there's a cleaner way to solve this but for the moment I made it work just deleting manually those assigned roles.

I was able to create a new chatbot with a separated index and sharing most of the infra, no matter it is located in other resource group.

Thank you again!
C.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants