🚀 DevHunter AI Dashboard

Automated Technical Recruitment Platform

Automated Screening • Deep Code Analysis • Voice AI Interviews

Report Bug · Request Feature

DevHunter AI is an automated technical recruitment platform that combines GitHub repository analysis with AI-driven voice interviews. It helps recruiters screen engineering candidates by deeply analyzing their code and conducting real-time, context-aware phone interviews.

DevHunter transforms the hiring pipeline by automating the initial technical screen. Instead of generic questions, it reads the candidate's actual code to generate highly specific, relevant interview topics.

• Deep GitHub Analysis: Scans repositories (including complex nested structures) to identify tech stacks, complexity metrics, and code quality indicators.
• Automated Voice Interviews: Leverages Telnyx and OpenAI to call candidates and ask technical questions generated specifically from their code analysis.
• Real-time Transcript: Watch the interview unfold live on the dashboard as the AI speaks with the candidate.
• Candidate Pipeline: Manage candidates through stages: New → Analyzing → Scheduled → Interviewing → Completed.
• Modern UI: Built with React, Tailwind CSS, and Shadcn UI for a professional SaaS aesthetic.

Before running the project, ensure you have the following installed and configured:

1. Node.js (v18+ recommended)
2. OpenAI API Key (for code analysis and interview logic)
3. Telnyx Account (API Key + Phone Number + Call Control App)
4. Ngrok (to expose your local server for voice webhooks)

Follow these steps to get a local development environment running.

git clone [https://github.com/your-username/devhunter-dashboard.git](https://github.com/your-username/devhunter-dashboard.git)
cd devhunter-dashboard

2. Install Dependencies Install the necessary packages for both the backend and frontend.

npm install

3. Configure Environment Variables Create a .env file in the root directory. You can copy the example below:

# Create the file
touch .env
Paste the following configuration into your .env file:

Code snippet

# Server Configuration
PORT=3000

# API Keys
OPENAI_API_KEY=sk-YOUR_OPENAI_KEY_HERE
TELNYX_API_KEY=KEY017_YOUR_TELNYX_KEY_HERE

# Voice Configuration
TELNYX_PHONE_NUMBER=+15551234567
# NOTE: This URL will be updated in Step 4
PUBLIC_URL=[https://placeholder.ngrok-free.app](https://placeholder.ngrok-free.app)

4. Start Ngrok Tunnel Telnyx needs to reach your local server to handle call events. Start Ngrok on your backend port:

ngrok http 3000

⚠️ Important: Copy the HTTPS URL provided by Ngrok (e.g., https://xyz.ngrok-free.app) and paste it into the PUBLIC_URL variable in your .env file.

5. Run the Application This command uses concurrently to start both the backend API (port 3000) and the frontend React app (port 5173).

npm start

6. Access the Dashboard Open your web browser and navigate to: http://localhost:5173

📖 Usage Guide Add a Candidate: Click the + icon in the sidebar. Enter a name, role, and a public GitHub repository URL.

Analyze Code: Select the candidate and click Analyze GitHub. The AI will deep-scan the repository.

Start Interview: Once analysis is complete, the status changes to "Scheduled." Click Start Interview, enter a phone number, and the AI will initiate the call.

Monitor Live: Watch the Live Transcript panel update in real-time as the candidate answers technical questions.

📂 Directory Structure Plaintext

devhunter-dashboard/
├── .env                    # Environment variables (API Keys)
├── package.json            # Project dependencies & scripts
├── server.js               # Node.js Backend (Voice & GitHub Logic)
├── vite.config.js          # Vite Configuration
├── tailwind.config.js      # Tailwind Configuration
├── postcss.config.js       # PostCSS Configuration
├── index.html              # React Entry Point
├── public/                 # Static assets
└── src/
    ├── App.jsx             # Main Application Component
    ├── main.jsx            # React DOM Render
    ├── index.css           # Global Tailwind Styles
    ├── components/         # Reusable UI components
    └── lib/                # Utility functions (e.g., API helpers)

##⚡ Troubleshooting Vite: react-babel error

Ensure you removed type imports from App.jsx.

Example: Change import { type ClassValue } ... → import { clsx } ...

Voice Call Not Connecting

Check your PUBLIC_URL in .env. It must match your current running Ngrok session. Also, ensure your Telnyx Call Control App is pointing to ${PUBLIC_URL}/webhooks/telnyx.

GitHub Analysis Fails

Ensure the repository is public. If hitting rate limits, try using a different repo or wait a few minutes.