🚀 About the Project 🔍 Inspiration We wanted to simplify resume screening and enhancement using AI—something lightweight, accessible, and private. Telegram felt like the perfect platform for fast interactions without needing a separate app or website. The goal was to let users upload any resume format and instantly gain structured insights, chat with AI, and even create beautifully designed PDFs.
🛠️ What We Built A Telegram bot with inline command support (/start, /convert, etc.)
AI-powered resume parsing using Google Gemini 2.0 Flash
Support for multiple formats: PDF, DOCX, TXT, and images
Chat-based AI interaction for resume analysis and interview preparation
HTML-to-PDF generation using Playwright and a clean HTML template
JSON-based data structure for dynamic content mapping in resumes
Full privacy: No personal data or resumes are stored beyond temporary processing
🧠 What We Learned Working with Playwright for headless PDF generation
Structuring AI prompts for accurate resume extraction
Handling file parsing for diverse resume formats
Integrating Telegram Bot API with seamless UX and inline keyboards
Building secure, ephemeral storage for sensitive data
⚔️ Challenges Faced Parsing inconsistent resume formats, especially scanned images
Maintaining structured JSON output across all file types
Setting up Playwright on Linux servers (like Render) with proper dependencies
Balancing user experience and privacy, ensuring files are deleted post-processing
Optimizing response speed without sacrificing accuracy
🔣 Tech Stack Python + telebot for the Telegram bot
Google Gemini 2.0 Flash for AI extraction
Playwright for HTML to PDF rendering
HTML/CSS template for PDF resume
OCR + File processing libraries (for image/TXT support)
Log in or sign up for Devpost to join the conversation.