This is a FastAPI app that provides an API for interacting with OpenAI models using Semantic Kernel SDK. It is meant to be used in conjunction with the store-admin app.
The app does not rely on any other services other than OpenAI or Azure OpenAI endpoints, so you can run it locally without any other services running.
To run the app, clone the repo, open a terminal, and navigate to the ai-service directory. Then run the following commands:
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
export USE_AZURE_OPENAI=True # set to False if you are not using Azure OpenAI
export USE_AZURE_AD=True # set to True if you are using Azure OpenAI with Azure AD authentication
export AZURE_OPENAI_API_VERSION=2024-02-15-preview # set to the version of the Azure OpenAI API you are using https://learn.microsoft.com/azure/ai-services/openai/reference#rest-api-versioning
export AZURE_OPENAI_DEPLOYMENT_NAME= # required if using Azure OpenAI
export AZURE_OPENAI_ENDPOINT= # required if using Azure OpenAI
export AZURE_OPENAI_DALLE_ENDPOINT= # required if using Azure OpenAI's DALL-E model
export AZURE_OPENAI_DALLE_DEPLOYMENT_NAME= # required if using Azure OpenAI's DALL-E model
export OPENAI_API_KEY= # always required if using OpenAI if using Azure OpenAI, consider use Workload Identity https://learn.microsoft.com/azure/aks/open-ai-secure-access-quickstart
export OPENAI_ORG_ID= # required if using OpenAI
uvicorn main:app --host 127.0.0.1 --port 5001When the app is running, you should see output similar to the following:
INFO: Started server process [134031]
INFO: Waiting for application startup.
INFO: Application startup complete.
INFO: Uvicorn running on http://127.0.0.1:5001 (Press CTRL+C to quit)
Using the test-ai-service.http file in the root of the repo, you can test the API. However, you will need to use VS Code and have the REST Client extension installed.