Skip to content

Latest commit

 

History

History
43 lines (28 loc) · 924 Bytes

File metadata and controls

43 lines (28 loc) · 924 Bytes

Tutor Questions Backend

This service provides the question-answering and evaluation engine for The Tutor platform.

Objective

  • Store, serve, and evaluate objective questions for students
  • Provide multi-agent, real-time feedback on student answers

Functionalities

  • CRUD for questions
  • Multi-agent evaluation via /grader/interaction
  • Submission history and answer storage

Infrastructure Requirements

  • Python 3.13+
  • FastAPI
  • Cosmos DB (Azure)
  • Azure OpenAI (for evaluation)

Running Locally

  1. Install dependencies:

    uv pip install --python .venv -e .[dev]
  2. Start the API:

    uvicorn app.main:app --reload

Deploying to Azure

  • Build a Docker image and push to ACR
  • Deploy as a container app using the azd + Terraform infrastructure
  • Configure environment variables for Cosmos DB and Azure OpenAI