Current Status: 🛠️ Development in progress. Testing frontend logic and PDF generation flow.
ResumeForge AI is an intelligent document generator that solves the "Blank Page Problem" for job seekers. Instead of worrying about formatting or phrasing, users simply input their raw, messy notes, and the AI architect builds a polished, professional PDF resume in seconds.
- 🧠 Intelligent Rewriting: Uses Google Gemini (via LangChain) to transform informal text (e.g., "I made a website") into professional "Action-Result" bullet points.
- 📄 PDF Engineering: Custom FPDF class generating pixel-perfect layouts with coordinate-based drawing.
- 🎨 Dynamic Theming: Switch between Modern (Blue/Bold) and Classic (Clean/Minimal) designs instantly.
- 🎯 Context Awareness: Paste a Target Job Description, and the AI will tailor your skills and summary to match the role keywords.
- 🛡️ Robust Text Sanitization: Automatically cleans and transcodes special characters to ensure PDF compatibility.
| Component | Technology | Description |
|---|---|---|
| Frontend | Streamlit | Interactive UI and State Management |
| LLM Orchestration | LangChain | Managing Prompts and Structured Output (JSON) |
| AI Model | Google Gemini 2.5 | The logic brain for rewriting content |
| Document Gen | FPDF | Low-level library for drawing the PDF canvas |
| Data Validation | Pydantic | Enforcing strict schema for AI responses |
This project is a deep dive into Backend GenAI Engineering.
While I architected the LangChain pipelines, Pydantic schemas, and Application Logic from scratch, I leveraged Google Gemini as a coding assistant to help generate the complex coordinate mathematics required for the FPDF frontend layout. This allowed me to focus 90% of my energy on the core AI logic and data flow, while treating the frontend code generation as an accelerated "drafting" process.
Follow these steps to set up the project on your machine:
git clone [https://github.com/your-username/ResumeForge-AI.git](https://github.com/your-username/ResumeForge-AI.git)
cd ResumeForge-AI
python -m venv venv
# On Windows:
venv\Scripts\activate
# On Mac/Linux:
source venv/bin/activate
pip install -r requirements.txt
Create a file named .env in the root directory and add your Google API key:
GOOGLE_API_KEY="your_actual_api_key_here"
streamlit run app.py
ResumeForge-AI/
├── app.py # Main Application (UI + PDF Generation Logic)
├── LangChainInput.py # Backend Logic (Gemini + Pydantic + Prompt)
├── requirements.txt # Project Dependencies
├── .env # API Keys (Not uploaded to GitHub)
├── .gitignore # Files to ignore (env, pycache)
├── LICENSE # MIT License
└── README.md # Documentation
Since this project is currently "Under Construction," I am primarily looking for feedback on the Resume Parsing logic. Feel free to open an issue if you find a bug!
Built with ❤️ by DEV DOSHI