Important: This application is currently in the testing phase. You might encounter bugs or issues while using it. We appreciate your understanding and patience.
ResumeParser is a powerful AI-driven tool that streamlines resume analysis by automatically extracting and analyzing key information from resumes. Built for recruiters and HR professionals, it combines advanced NLP with efficient processing to deliver accurate insights.
Core Capabilities
-
📄 Intelligent parsing for PDFs with mutliple LLM model support
- Hermes LLama3.1 8B - 8 Bit quantized model
- Hermes LLama3.2 3B - 8 Bit quantized model
- IBM Granite 3.1 8B - 8 Bit quantized model
-
🔍 Missing section identification
-
✍️ Advanced spell checking analysis
-
💡 AI-powered interview question generation
-
📊 Visual resume analysis with WordCloud generation
 |
| ResumeParser LIGHT |
| Category | Technologies |
|---|---|
| Backend | FastAPI, Python 3.8+ |
| ML/AI | PyTorch, Transformers |
| UI | Gradio |
| Document Processing | IBM Docling |
| Local LLM Software | LM Studio |
| LLM Models | Llama 3.x, IBM Granite |
Before running the application, ensure your system meets the following requirements:
- Python: Version 3.10 or higher
- RAM: Minimum 8GB (16GB recommended)
- GPU: Recommended for optimal performance (NVIDIA RTX 3050 or higher)
- RAM: 16 GB DDR5
- GPU: NVIDIA RTX 3060 (6GB)
- Processor: Intel i7-12700H
python -m venv venv
source venv/bin/activate # Windows: venv\Scripts\activate# clone repo
git clone https://github.com/SID-SURANGE/ResumeParser.git
cd ResumeParser
# Install dependencies
pip install -r requirements.txtUse the below command to store the HuggingFace API key and other API keys for future integrations.(Optional for current setup)
cp .env.example .env
After installation:
- Download preferred models from LM Studio's built-in model hub.
- Currently the tested models are Hermes LLama3.1 8B, Hermes LLama3.2 3B, IBM Granite 3.1 8B - 8 Bit quantized model
- Load and start the local server for model inference
- Keep the server running while using ResumeParser
Watch step-by-step tutorial on Youtube by Developers Digest:
Once LM Studio is configured and its server is started up, follow below steps -
# development mode,
uvicorn main:app --reload
# production mode
uvicorn main:app --host 0.0.0.0 --port 8000 --reload- Access the web interface at http://localhost:8000
- Theme can be modified by appending to url - http://127.0.0.1:8000/?__theme=light or http://127.0.0.1:8000/?__theme=dark
- Upload a PDF resume (works best for 1 page resumes)
- Select an LLM model
- Click "Parse" to analyze the resume
- Use additional features (enabled only after Parse button click):
- 🤔 Generate interview questions
- 📊 Create word clouds
| Category | Endpoints | Description |
|---|---|---|
| Resume parsing |
/api/v1/parse | Parse Uploaded resume |
| Question generation |
/api/v1/questions | Generate interview questions |
API documentation is available at http://localhost:8000/docs.
 |
 |
| Input Resume (Overleaf Template) | Parser Output |
We are actively working on adding new features:
- 💼 Intelligent job role matching.
- 📄 Multi-page resume support.
- 🤖 Integration with advanced models like Claude and OpenAI.
- 🖥️ Model Loading Errors: Ensure GPU requirements are met.
- 📁 File Processing Issues: Verify that only PDF files are uploaded.
- 💾 Memory Errors: Increase system RAM or optimize workloads.
- 🔍 Error messages need better user clarity.
- 🛠️ Handling unexpected model responses.
- 📑 Multi-page document support is under development.
- 📊 Complex document handling (e.g., tables, graphics) is being improved.
Licensed under MIT License. See LICENSE for details.
For any issues or feature requests:
- 🐛 Report bugs via GitHub Issues.
Made with ❤️ by the ResumeParser Team
If you find this project useful, please consider giving it a star ⭐ and forking it 🍴. Your support is greatly appreciated!