InterviewAI Backend

AI-powered mock interview API built with FastAPI and OpenAI GPT-4. This service generates tailored interview questions based on a candidate's profile and provides real-time conversational interviews with AI-driven feedback.

Live API: interviewai-backend-z8rg.onrender.com API Docs: interviewai-backend-z8rg.onrender.com/docs Frontend Repo: InterviewAI-Frontend

---

Architecture

Streamlit Frontend ──► FastAPI Backend ──► OpenAI GPT-4
                            │
                        SQLAlchemy
                            │
                         SQLite DB

The backend handles all AI logic and session management, keeping the OpenAI API key secure on the server side. The frontend communicates exclusively through REST endpoints.

---

Tech Stack

Layer	Technology
API Framework	FastAPI
AI/LLM	OpenAI GPT-4
Database	SQLite + SQLAlchemy
Containerization	Docker
Deployment	Render (Docker Web Service)

---

Project Structure

InterviewAI-Backend/
├── 7_final.py           # Main FastAPI application (routes, session logic, OpenAI calls)
├── database.py          # SQLAlchemy models and database connection
├── llm.py               # LLM prompt construction and OpenAI API wrapper
├── requirements.txt     # Python dependencies
├── Dockerfile           # Production Docker image
├── docker-compose.yml   # Local development setup
└── .gitignore

---

API Endpoints

Health Check

GET /

Returns server status and active session count.

{
  "status": "running",
  "active_sessions": 0
}

Start Interview

POST /start_interview

Request Body:

{
  "name": "Ramu",
  "experience": "3 years ML",
  "skills": "Python, ML, Deep Learning",
  "position": "Data Scientist",
  "company": "Amazon"
}

Response: AI-generated interview question tailored to the candidate's profile, role, and target company.

Continue Interview

Submit answers and receive follow-up questions throughout the interview session.

Get Feedback

After completing the interview, receive a performance summary with scoring and improvement suggestions.

Full interactive docs available at /docs (Swagger UI) when the server is running.

---

Getting Started

Prerequisites

• Python 3.10+
• OpenAI API key

Local Setup

1. Clone the repo:

git clone https://github.com/RamuGanta/InterviewAI-Backend.git
cd InterviewAI-Backend

2. Install dependencies:

pip install -r requirements.txt

3. Set environment variables:

Create a .env file in the project root:

OPENAI_API_KEY=your_openai_api_key_here

4. Run the server:

uvicorn 7_final:app --host 0.0.0.0 --port 8000 --reload

The API will be available at http://localhost:8000 and docs at http://localhost:8000/docs.

Docker

docker build -t interviewai-backend .
docker run -p 8000:8000 --env-file .env interviewai-backend

Or with Docker Compose:

docker-compose up

---

Deployment

Deployed on Render as a Docker Web Service.

• Service type: Docker Web Service
• Health check: GET /
• Environment variables: OPENAI_API_KEY is stored securely in Render's environment settings (never committed to source control)

CORS Configuration

The backend allows cross-origin requests from the Streamlit frontend:

app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],
    allow_methods=["*"],
    allow_headers=["*"],
)

---

Security

• API keys are stored only in environment variables, never in source code
• .env is included in .gitignore
• GitHub push protection is enabled to prevent accidental secret exposure
• All OpenAI calls are proxied through the backend — the frontend never touches the API key

---

Related

• Frontend: InterviewAI-Frontend — Streamlit UI for the interview experience
• Earlier Prototype: Interview-and-Feedback-tool — Single-file Streamlit + OpenAI version

---

Author

Ramu Ganta — LinkedIn · GitHub