Knitty: The Resume Factory

Creative Geek•

Knitty is an intelligent CV tailoring system that optimizes resumes for specific job postings using LLMs and cosine similarity analysis.

Knitty: The Resume Factory

🚀 Overview

Knitty is an AI-powered CV tailoring system that optimizes resumes for specific job postings. The system analyzes your existing CV alongside a target job posting, then generates an optimized version that better aligns with the job requirements while maintaining authenticity and professional quality.

✨ Key Features

  • 🔄 Multi-LLM Architecture: Leverages four specialized AI models for different tasks
  • 📄 PDF Processing: Extracts and processes CV content from PDF files
  • 🌐 Web Scraping: Automatically fetches job postings from LinkedIn and other platforms
  • 🧠 RAG-Powered Analysis: Uses Retrieval-Augmented Generation for intelligent job content extraction
  • 📊 Similarity Scoring: Measures CV-job alignment using cosine similarity
  • 🎯 Smart Enhancement: Generates tailored CVs while maintaining professional quality
  • 🔁 Iterative Improvement: Automatically retries generation if similarity doesn't improve

🏗️ Architecture

Architecture Diagram

The system uses Multi-LLM Architecture:

  • Fast LLM (for example: GPT-OSS-120B): Quick keyword extraction from CV and job postings
  • Context LLM: Large-context processing for web-scraped content (deprecated, was used in No-RAG mode)
  • Embed LLM: Generates embeddings for similarity analysis
  • Smart LLM (Currently recommending Gemini 2.5 Pro): CV enhancement and generation

To get started just follow the repo here: https://github.com/Creative-Geek/Knitty

Protip: You can use GimmeCV to render the output in a compact single page ATS compliant pdf.

For any questions or issues, You can directly contact me on LinkedIn: https://www.linkedin.com/in/ahmed-taha-thecg/