Welcome to the Azure AI Assistants directory. This pillar is focused on task-specific assistive AI implementations primarily utilizing the structure of the Azure OpenAI Assistants API.
This folder contains the following implementations:
Labs/: Step-by-step labs guiding you through the creation of your first Assistant, demonstrating how to assign instructions, equip the Assistant with tools, and run conversational Threads.
For the Python implementations in this directory:
-
Activate your Virtual Environment:
# From the repository root source venv/bin/activate
(Windows:
venv\Scripts\activate) -
Environment Variables: The Assistants API requires your Azure OpenAI endpoint and key. Always configure your local
.envvia the provided.env.exampleblueprint located in the lab folder.
- File Search Data: If demonstrating the "File Search" capability of the Assistants API, ensure you only use non-sensitive, mock data (like sample manuals or generic markdown files). Do not commit proprietary documents.
- Commiting: Prefix your commits with
feat(assist):orfix(assist):. - Pushing: Before pushing, ensure that your scripts clean up the Assistants, Threads, and Vector Stores they dynamically create on Azure to prevent resource leaks and extra billing on the shared dev accounts.