{"id":33329355,"url":"https://github.com/pratulparmar/interview-ace-ai","last_synced_at":"2026-04-09T22:32:02.545Z","repository":{"id":324020633,"uuid":"1095491229","full_name":"pratulparmar/interview-ace-ai","owner":"pratulparmar","description":"AI-powered technical interview assistant using RAG, Multi-Agent Systems, LangChain, and GPT-4","archived":false,"fork":false,"pushed_at":"2025-11-13T10:21:49.000Z","size":652,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2025-11-13T12:14:00.123Z","etag":null,"topics":["artificial-intelligence","faiss","gpt-4","langchain","machine-learning","multiagent-systems","openai","python","rag","streamlit","vector-database"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/pratulparmar.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2025-11-13T05:50:59.000Z","updated_at":"2025-11-13T10:21:54.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/pratulparmar/interview-ace-ai","commit_stats":null,"previous_names":["pratulparmar/interview-ace-ai"],"tags_count":null,"template":false,"template_full_name":null,"purl":"pkg:github/pratulparmar/interview-ace-ai","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pratulparmar%2Finterview-ace-ai","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pratulparmar%2Finterview-ace-ai/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pratulparmar%2Finterview-ace-ai/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pratulparmar%2Finterview-ace-ai/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/pratulparmar","download_url":"https://codeload.github.com/pratulparmar/interview-ace-ai/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pratulparmar%2Finterview-ace-ai/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":285465933,"owners_count":27176533,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","status":"online","status_checked_at":"2025-11-20T02:00:05.334Z","response_time":54,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["artificial-intelligence","faiss","gpt-4","langchain","machine-learning","multiagent-systems","openai","python","rag","streamlit","vector-database"],"created_at":"2025-11-20T16:02:12.860Z","updated_at":"2025-11-20T16:03:37.881Z","avatar_url":"https://github.com/pratulparmar.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# 🎯 InterviewAce AI\n\n**AI-Powered Technical Interview Assistant with RAG, Multi-Agent System, and LangChain**\n\n[![Python 3.9+](https://img.shields.io/badge/python-3.9+-blue.svg)](https://www.python.org/downloads/)\n[![LangChain](https://img.shields.io/badge/LangChain-Latest-green.svg)](https://python.langchain.com/)\n[![OpenAI](https://img.shields.io/badge/OpenAI-GPT--4-orange.svg)](https://openai.com/)\n[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)\n\nA production-ready GenAI application that conducts technical interviews, evaluates answers using AI agents, and provides comprehensive feedback with detailed reports.\n\n---\n\n## 🌟 Features\n\n### Core Capabilities\n- **🔍 RAG-Powered Question Retrieval** - Semantic search through 40+ curated interview questions\n- **🤖 AI Question Generation** - GPT-4 generates custom questions based on job descriptions\n- **🎤 Interactive Interviews** - Conduct full technical interviews via CLI or Web UI\n- **📊 AI Evaluation** - Multi-criteria scoring with detailed feedback\n- **📈 Comprehensive Reports** - Overall scores, category breakdown, and actionable insights\n\n### Technical Highlights\n- **Multi-Agent System** - Specialized agents for generation, interviewing, and evaluation\n- **Vector Database** - FAISS for fast semantic search\n- **LangChain Integration** - Production-ready LLM orchestration\n- **Prompt Engineering** - Structured outputs with rubric-based evaluation\n- **Hybrid Approach** - RAG retrieval + AI generation for optimal results\n\n---\n\n## 🏗️ Architecture\n\n```\n┌─────────────────────────────────────────────────────────────┐\n│                    InterviewAce AI System                    │\n├─────────────────────────────────────────────────────────────┤\n│                                                               │\n│  ┌──────────────┐    ┌──────────────┐    ┌──────────────┐  │\n│  │ Job Description │──→│ RAG Retriever │──→│  Questions   │  │\n│  └──────────────┘    └──────────────┘    └──────────────┘  │\n│         │                     │                               │\n│         │             ┌───────────────┐                       │\n│         └────────────→│   Generator   │                       │\n│                       │     Agent     │                       │\n│                       └───────────────┘                       │\n│                               │                               │\n│                               ↓                               │\n│                       ┌───────────────┐                       │\n│                       │  Interviewer  │                       │\n│                       │     Agent     │                       │\n│                       └───────────────┘                       │\n│                               │                               │\n│                               ↓                               │\n│                       ┌───────────────┐                       │\n│                       │   Evaluator   │                       │\n│                       │     Agent     │                       │\n│                       └───────────────┘                       │\n│                               │                               │\n│                               ↓                               │\n│                       ┌───────────────┐                       │\n│                       │  Final Report │                       │\n│                       └───────────────┘                       │\n│                                                               │\n└─────────────────────────────────────────────────────────────┘\n```\n\n### Component Overview\n\n| Component | Technology | Purpose |\n|-----------|-----------|---------|\n| **RAG Retriever** | FAISS + OpenAI Embeddings | Semantic search through question bank |\n| **Question Generator** | GPT-4 + LangChain | Generate custom interview questions |\n| **Interviewer Agent** | GPT-4 + LangChain | Conduct natural conversations |\n| **Evaluator Agent** | GPT-4 + Structured Outputs | Score answers with detailed feedback |\n| **Vector Store** | FAISS | Store and retrieve 3072-dim embeddings |\n| **Orchestrator** | Python | Coordinate all components |\n\n---\n\n## 🚀 Quick Start\n\n### Prerequisites\n\n- Python 3.9 or higher\n- OpenAI API key\n- Git\n\n### Installation\n\n1. **Clone the repository**\n   ```bash\n   git clone https://github.com/yourusername/interview-ace-ai.git\n   cd interview-ace-ai\n   ```\n\n2. **Create virtual environment**\n   ```bash\n   python -m venv venv\n   \n   # Windows\n   venv\\Scripts\\activate\n   \n   # Mac/Linux\n   source venv/bin/activate\n   ```\n\n3. **Install dependencies**\n   ```bash\n   pip install -r requirements.txt\n   ```\n\n4. **Set up environment variables**\n   ```bash\n   # Create .env file\n   cp .env.example .env\n   \n   # Edit .env and add your OpenAI API key\n   # OPENAI_API_KEY=your_api_key_here\n   ```\n\n5. **Initialize question bank**\n   ```bash\n   python backend/scripts/question_bank_initializer.py\n   ```\n   This creates a vector database with 40+ curated questions (takes ~1 minute).\n\n---\n\n## 💻 Usage\n\n### Option 1: Interactive CLI\n\nRun an interview in your terminal:\n\n```bash\npython cli_interview.py\n```\n\n**Features:**\n- Interactive Q\u0026A\n- Real-time answer submission\n- Immediate evaluation\n- Detailed feedback\n\n**Demo:**\n```\n🎯 INTERVIEW ACE AI - Interactive Technical Interview\n\nEnter your name: John Doe\nJob Title: Senior Python Engineer\n\n[Paste job description]\n[Answer 5 questions]\n[Receive comprehensive evaluation]\n```\n\n### Option 2: Streamlit Web UI\n\nLaunch the web interface:\n\n```bash\nstreamlit run app.py\n```\n\nThen open your browser to `http://localhost:8501`\n\n**Features:**\n- Beautiful, intuitive interface\n- Progress tracking\n- Question-by-question navigation\n- Visual score breakdown\n- Download reports (JSON/TXT)\n\n### Option 3: Programmatic API\n\nUse the orchestrator directly in your code:\n\n```python\nfrom backend.orchestrator.interview_orchestrator import InterviewOrchestrator\n\n# Initialize\norchestrator = InterviewOrchestrator()\n\n# Run complete interview\nsession = orchestrator.run_complete_interview(\n    candidate_name=\"Alice Smith\",\n    job_description=\"Senior Python Engineer with ML experience...\",\n    job_title=\"Senior Python Engineer\",\n    num_questions=5,\n    simulate=False,  # Set to True for testing\n    use_rag=True\n)\n\n# Get report\nprint(session.detailed_feedback)\n\n# Save session\norchestrator.save_session(session)\n```\n\n---\n\n## 📊 Sample Output\n\n```\n================================================================================\nINTERVIEW REPORT\n================================================================================\n\nCANDIDATE INFORMATION\n---------------------\nName: John Doe\nPosition: Senior Python Engineer\nInterview Date: 2025-11-08T...\n\nOVERALL PERFORMANCE\n-------------------\nOverall Score: 7.5/10\nRecommendation: Hire - Solid performance with minor gaps\n\nCATEGORY BREAKDOWN\n------------------\nPython......................... 8.0/10\nLLMs........................... 7.5/10\nRAG............................ 7.0/10\nLangChain...................... 7.5/10\n\nQUESTION-BY-QUESTION ANALYSIS\n------------------------------\n\nQuestion 1: What is RAG and why is it useful?\nScore: 7.5/10\n  Technical Accuracy: 8/10\n  Completeness: 7/10\n  Communication: 8/10\n  Depth: 7/10\n\n  Strengths:\n    ✓ Clear explanation of RAG fundamentals\n    ✓ Mentioned key use cases\n\n  Areas for Improvement:\n    → Could discuss implementation details\n    → Add more real-world examples\n\n[... more questions ...]\n\nFINAL RECOMMENDATION\n--------------------\nHire - Solid performance with minor gaps\n================================================================================\n```\n\n---\n\n## 🗂️ Project Structure\n\n```\ninterview-ace-ai/\n├── backend/\n│   ├── agents/\n│   │   ├── question_generator.py    # GPT-4 question generation\n│   │   ├── interviewer_agent.py     # Conversation management\n│   │   └── evaluator_agent.py       # Answer scoring\n│   ├── rag/\n│   │   ├── embeddings.py            # OpenAI embeddings wrapper\n│   │   ├── vector_store.py          # FAISS vector database\n│   │   └── rag_retriever.py         # Semantic search logic\n│   ├── orchestrator/\n│   │   └── interview_orchestrator.py # Main pipeline coordinator\n│   ├── config/\n│   │   └── settings.py              # Configuration management\n│   ├── scripts/\n│   │   └── question_bank_initializer.py # Populate vector DB\n│   └── data/\n│       ├── vector_store/            # FAISS index + metadata\n│       └── interviews/              # Saved interview sessions\n├── cli_interview.py                 # Interactive CLI application\n├── app.py                           # Streamlit web UI\n├── requirements.txt                 # Python dependencies\n├── .env.example                     # Environment template\n└── README.md                        # This file\n```\n\n---\n\n## 🔧 Configuration\n\n### Environment Variables\n\nCreate a `.env` file with:\n\n```env\n# OpenAI API Configuration\nOPENAI_API_KEY=your_openai_api_key_here\n\n# Optional: LangSmith Tracing (for debugging)\nLANGCHAIN_TRACING_V2=true\nLANGCHAIN_API_KEY=your_langsmith_key_here\nLANGCHAIN_PROJECT=interview-ace-ai\n```\n\n### Customization\n\n**Add more questions:**\n```python\n# Edit backend/scripts/question_bank_initializer.py\n# Add your questions to the appropriate category function\n```\n\n**Adjust evaluation criteria:**\n```python\n# Edit backend/agents/evaluator_agent.py\n# Modify the rubric in the prompt\n```\n\n**Change models:**\n```python\n# Edit backend/config/settings.py\n# Modify model settings (e.g., use gpt-3.5-turbo for cost savings)\n```\n\n---\n\n## 🧪 Testing\n\n### Test Individual Components\n\n**Test embeddings:**\n```bash\npython backend/rag/embeddings.py\n```\n\n**Test vector store:**\n```bash\npython backend/rag/vector_store.py\n```\n\n**Test RAG retriever:**\n```bash\npython backend/rag/rag_retriever.py\n```\n\n**Test complete pipeline:**\n```bash\npython backend/orchestrator/interview_orchestrator.py\n```\n\n### Run Full System Test\n\n```bash\n# With simulated answers\npython backend/orchestrator/interview_orchestrator.py\n\n# Interactive test\npython cli_interview.py\n```\n\n---\n\n## 💰 Cost Estimation\n\n| Operation | Cost per 1000 tokens | Typical Interview Cost |\n|-----------|---------------------|------------------------|\n| Embeddings (3072-dim) | $0.00013 | ~$0.05 |\n| GPT-4 Question Gen | $0.01 input / $0.03 output | ~$0.15 |\n| GPT-4 Evaluation | $0.01 input / $0.03 output | ~$0.50 |\n| **Total per 5-question interview** | | **~$0.70** |\n\n**Tips to reduce costs:**\n- Use RAG retrieval (avoids generation cost)\n- Switch to GPT-3.5-turbo for evaluation (~10x cheaper)\n- Batch process multiple interviews\n- Cache embeddings (already implemented)\n\n---\n\n## 🎓 What This Project Demonstrates\n\n### GenAI Engineering Skills\n✅ **RAG Implementation** - Vector DB, semantic search, retrieval strategies  \n✅ **Multi-Agent Systems** - Specialized agents with distinct roles  \n✅ **LLM Orchestration** - LangChain chains, prompts, and structured outputs  \n✅ **Prompt Engineering** - Few-shot learning, CoT, structured JSON  \n✅ **Vector Databases** - FAISS integration and optimization  \n✅ **Production Architecture** - Modular, scalable, testable design  \n\n### Software Engineering Skills\n✅ **Clean Code** - Type hints, docstrings, separation of concerns  \n✅ **System Design** - Component-based architecture  \n✅ **Error Handling** - Graceful failures and user feedback  \n✅ **Testing** - Unit tests for each component  \n✅ **Documentation** - Comprehensive README and code comments  \n✅ **UX Design** - Both CLI and Web interfaces  \n\n---\n\n## 🛠️ Technologies Used\n\n- **LLMs:** OpenAI GPT-4, GPT-3.5-turbo\n- **Embeddings:** OpenAI text-embedding-3-large (3072 dimensions)\n- **Vector DB:** FAISS (Facebook AI Similarity Search)\n- **Framework:** LangChain, LangGraph\n- **Web UI:** Streamlit\n- **Language:** Python 3.9+\n- **Monitoring:** LangSmith (optional)\n\n---\n\n## 🤝 Contributing\n\nContributions are welcome! Please feel free to submit a Pull Request.\n\n### Areas for Enhancement\n- [ ] Add more question categories (System Design, Behavioral, etc.)\n- [ ] Implement voice interview mode (speech-to-text)\n- [ ] Add multi-language support\n- [ ] Create analytics dashboard\n- [ ] Implement user authentication\n- [ ] Add interview recording/replay\n- [ ] Build mobile app version\n\n---\n\n## 📝 License\n\nThis project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.\n\n---\n\n## 🙏 Acknowledgments\n\n- OpenAI for GPT-4 and embeddings\n- LangChain team for the excellent framework\n- Facebook Research for FAISS\n- Streamlit for the web framework\n\n---\n\n## 📧 Contact\n\n**Your Name**  \n- GitHub: [@pratulparmar](https://github.com/pratulparmar)\n- LinkedIn: [Pratul Parmar](https://www.linkedin.com/in/pratul-parmar-a5002417a?lipi=urn%3Ali%3Apage%3Ad_flagship3_profile_view_base_contact_details%3BOW%2B1F9BwTm6kxUwnRt92yg%3D%3D)\n- Email: pratulparmar8@gmail.com\n\n---\n\n## ⭐ Star this repo\n\nIf you find this project helpful, please give it a star! It helps others discover this work.\n\n---\n\n**Built with ❤️ as a portfolio project demonstrating GenAI engineering skills**","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fpratulparmar%2Finterview-ace-ai","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fpratulparmar%2Finterview-ace-ai","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fpratulparmar%2Finterview-ace-ai/lists"}