Building an Open-Source AI Mock Interviewer with Resume Personalization & Answer Coaching published: true

🎃 Hacktoberfest 2026: The Open-Source AI Challenge
Preparing for high-stakes interviews is often stressful. Static question lists don't recreate the pressure of dynamic conversations, nor do they probe into a candidate's actual resume claims.

For Week 1 of the Hacktoberfest Open-Source AI Challenge, I built AI Interview Simulator — an intelligent pair interviewer powered entirely by open-source LLMs.

🚀 Live Demo & Links

• 🌐 Live Demo (Frontend): https://ai-interview-simulator-nu-weld.vercel.app/
• ⚡ Backend API (Render): https://ai-interview-simulator-9v89.onrender.com/

- 💻 GitHub Repository: https://github.com/sage106/ai-interview-simulator

💡 Key Features

1. 📄 Resume-Driven Question Generation: Upload your resume (PDF/TXT) or paste your experience. The AI analyzes your projects, tech stack, and career timeline to generate customized questions targeting your background.
2. 🎯 Multi-Industry Support: Pre-configured paths across Tech, AI & Data Science, Product, Design, Marketing & Sales, Finance, and Healthcare (plus custom career write-in support).
3. 💡 "How to Improve Your Answer" Coaching: Instead of just a numerical score, candidates receive concrete coaching on how to elevate their response (structure with STAR method, cite trade-offs, metrics, edge cases).
4. ✨ Ideal / Model Answers: An interactive reveal toggle showing the exact industry-standard response a senior hiring manager looks for.
5. 📈 Comprehensive Evaluation: Circular score meter, hiring decision recommendations (Strongly Recommend, Recommend, Needs Work), top strengths, and areas to improve.

6. 💾 MongoDB Session History: Stores previous sessions to track performance over time.

🛠️ Tech Stack

• Frontend: React 18, Vite, Tailwind CSS, Lucide Icons (deployed on Vercel)
• Backend: Node.js, Express, Multer, pdf-parse (deployed on Render)
• Database: MongoDB Atlas

- AI Core: Open-Source LLMs (via high-speed Groq API)

🧠 Architecture Flow

text
Candidate Resume (PDF/TXT) ──> pdf-parse extraction
                                   │
                                   ▼
Open-Source LLM (Groq) ──────> Dynamic Role/Resume Question ──> Candidate Response
                                   │
                                   ▼
Open-Source LLM (Groq) ──────> Dual Evaluation:
                               ├─ Score & Assessment Feedback
                               ├─ "How to Improve" Action Plan
                               └─ Model / Correct Answer
                                   │
                                   ▼
Session Completed ───────────> Comprehensive Report & MongoDB Session Storage

Comments

[DEV SUPPORTS]

You need to verify your account.

tr.ee/dev-to