AI Technical Interview System

📌 Project Overview

AI Technical Interview System is a standalone AI-powered technical screening platform designed to automate candidate interviews through conversational assessments, voice interaction, answer evaluation, and real-time candidate monitoring.

The platform conducts structured technical interviews, evaluates responses using NLP-based similarity scoring, tracks candidate behavior through browser telemetry and face monitoring, and generates detailed performance reports for recruiters.

---

Features

AI-Powered Technical Interviews

• ✅ Automated technical screening workflows
• ✅ Structured Python and SQL interview question banks
• ✅ Stateful interview progression and session management
• ✅ Dynamic question delivery and answer recording
• ✅ AI-assisted conversational interview experience

Candidate Evaluation

• ✅ TF-IDF based answer scoring
• ✅ Keyword matching and response relevance analysis
• ✅ Performance report generation
• ✅ Strength and weakness identification
• ✅ Detailed question-by-question evaluation

Interview Monitoring

• ✅ Real-time face tracking
• ✅ Face absence detection
• ✅ Browser tab-switch detection
• ✅ Candidate activity monitoring
• ✅ Interview integrity enforcement

Voice & Interaction

• ✅ Neural text-to-speech responses
• ✅ Voice-guided interview flow
• ✅ Audio response generation

Session Management

• ✅ Session persistence and recovery
• ✅ Local interview state caching
• ✅ Resume interrupted interviews
• ✅ Stateful candidate tracking

---

Technology Stack

Backend

• Python
• Flask

Artificial Intelligence

• Google Gemini API
• TF-IDF Vectorization
• Cosine Similarity Scoring

Frontend AI Models

• TensorFlow.js
• Tiny Face Detector
• COCO-SSD Object Detection

Database

• MongoDB

Voice Processing

• Edge TTS

---

Project Structure

AI_Interview_Standalone/
│
├── app.py
├── evaluator.py
├── interview_engine.py
├── questions.py
├── state_manager.py
├── requirements.txt
├── start.bat
├── .env
├── .gitignore
│
├── static/
│   ├── audio/
│   └── models/
│       ├── tiny_face_detector_model-shard1
│       ├── tiny_face_detector_model-weights_manifest.json
│       └── coco-ssd/
│
├── templates/
│   ├── index.html
│   ├── interview.html
│   └── result.html
│
└── README.md

---

Installation & Setup

1️⃣ Clone the Repository

git clone https://github.com/Rubel286/AI-Technical-Interview-System.git
cd AI_Interview_Standalone

2️⃣ Create & Activate Virtual Environment

Windows

python -m venv venv
venv\Scripts\activate

macOS / Linux

python3 -m venv venv
source venv/bin/activate

3️⃣ Install Dependencies

pip install -r requirements.txt

4️⃣ Configure Environment Variables

Create a .env file in the project root:

SECRET_KEY="your key"

MONGO_URI=mongodb://127.0.0.1:27017/recruitmentDB

GEMINI_API_KEY="your api key"

---

Run the Application

Verify MongoDB Connection

mongosh

Start the Server

python app.py

Or use the Windows launcher:

start.bat

---

Access the Application

Open your browser and navigate to:

http://127.0.0.1:5050

---

Integration with PeopleOps

This service can operate independently or integrate directly with the PeopleOps recruitment platform.

Workflow

1. Recruiter creates an interview request inside PeopleOps.
2. A secure interview token is generated.
3. The candidate receives a personalized interview link.
4. The candidate completes the AI-powered screening session.
5. Interview results are automatically stored in MongoDB.
6. Recruiters review performance reports inside the recruitment platform.

Stored Evaluation Data

The service generates:

• Interview Score
• Question-by-Question Evaluation
• Candidate Strengths
• Candidate Weaknesses
• Detailed Interview Reports
• Interview Completion Status

---

Screenshots

Landing Page

Interview Portal

Result Dashboard

---

Important

This project uses Google Gemini API.

Get your API key from:

https://aistudio.google.com/app/apikey

---

License

This project is licensed under the MIT License.

See the LICENSE file for details.