💼 Career OS — Resume AI Builder

Where Your Career Story Meets AI Intelligence

---

📖 Overview

Career OS — Resume AI Builder is a personal career intelligence hub and full-stack assistant. It extracts unstructured career data from uploaded resumes or external sites, organizes it into a centralized schema, and generates tailored resumes, cover letters, and web portfolios using state-of-the-art AI.

Purpose

• Automate the parsing of various resume formats (including scanned PDFs/images) into clean structured data.
• De-duplicate and sync career records across multiple source uploads automatically.
• Formulate custom-tailored, editable DOCX and printer-friendly HTML resumes optimized for specific job requirements.
• Quantify applicant-to-job compatibility scores, highlighting skill gaps and suggesting contextually relevant achievements.

---

✨ Features

🚀 Core Experience

• Scanned document parsing using high-accuracy Tesseract OCR with automatic image preprocessing.
• Multi-source de-duplication to seamlessly merge data across different documents.
• Interactive, schema-enforced profile manager for standard and custom career sections.
• High-fidelity, direct python-docx compilation for robust, editable Word document output.
• Provider-agnostic AI connection supporting both Google Gemini and OpenAI.

---

🎨 Interactive Sections

Section	Description	Tech
Dashboard	High-level profile health, upload statistics, and quick-export templates.	React, Zustand, Lucide
Profile Editor	Multi-module workspace for editing experience, projects, skills, and awards.	React, Tailwind CSS
Resume Builder	Custom styling engine for layouts, colors, and margins with instant code preview.	React, Jinja2, docx
Cover Letter	Automated cover letter writer tailored by industry, target company, and custom tone.	React, Gemini API
Portfolio	Fully responsive portfolio website generator with pre-built themes.	React, Tailwind, Gemini
Job Matcher	Deep alignment checker offering compatibility percentages and gap analysis.	React, FastAPI, ChromaDB

---

💻 Snippets Hub

• Multi-format document compilation engine supporting PDF, DOCX, and clean Markdown exports.
• Custom prompt injection preventing hallucinated achievements or fabricated skills.
• Tailored styling profiles matching precise font sizing, spacing, borders, and margins.
• Dynamic skill categorization separating expert-level competencies from secondary libraries.
• Direct code generation workspace displaying clean HTML/CSS outputs.

Categories

• Professional (Two-column, ATS compatible, blue accents)
• Creative (Bold typography, standout accent highlights, graphical elements)
• Minimalist (Traditional serif, elegant spacing, ultra-clean)
• Executive (Highly structured, rich meta descriptors, conservative margins)
• Single-Page Portfolio Themes
• Personalized Application Letters

---

🎮 Playground

• Live HTML code viewer and editor interface.
• Instant visual refreshing when custom templates or accent colors change.
• Raw schema inspection displaying structured profile JSON.
• Fast copy-to-clipboard functionality.

Shortcuts

Shortcut	Action
Ctrl + S	Export to DOCX / Save Profile
Ctrl + R	Re-generate with AI
Ctrl + P	Open Print/PDF Interface

---

🎨 Theme System

• Presets for Resume layout (Professional Blue, Slate, Emerald, Royal Purple, Accent Orange).
• Portfolio presets (Modern Dark with glassmorphism, Minimalist Light, Energetic Bold, Monospace Retro Terminal).
• Persistent user styling preferences stored in relational database.
• Dynamic local theme switching with minimal client-side performance cost.

---

🛠️ Tech Stack

Core

• React 19
• TypeScript 5
• Vite 6
• FastAPI 0.115
• SQLAlchemy 2.0 (Async)

Styling & Animation

• Tailwind CSS v4
• Lucide React Icons

3D

• Not utilized for core layouts; focused on lightweight vector styling and custom 3D CSS transform depth effects for responsive portfolios.

State

• Zustand
• Pydantic v2 (Data Validation)

Tools

• Google Generative AI / OpenAI API
• python-docx (Word builder)
• Tesseract OCR & OpenCV
• BeautifulSoup4 & lxml
• ChromaDB
• WeasyPrint & Jinja2

---

📦 Installation

Requirements

• Node.js 20+
• Python 3.11+
• PostgreSQL 16+
• Tesseract OCR (Optional, required for scanned resume OCR processing)

Setup

# Clone the repository
git clone https://github.com/eshitha06/resume_ai_builder.git
cd resume_ai_builder

# 1. Start the Backend
cd backend
python -m venv venv

# Windows
.\venv\Scripts\activate
# macOS/Linux
source venv/bin/activate

pip install -r requirements.txt
cp .env.example .env
python -m alembic upgrade head
python -m uvicorn app.main:app --reload

# 2. Start the Frontend (in a separate terminal)
cd ../frontend
npm install
cp .env.example .env.local
npm run dev

Build

# Production frontend build
cd frontend
npm run build
npm run preview

---

⚡ Performance

Optimizations

• Non-LLM programmatic DOCX generation executing in milliseconds.
• Asynchronous database access using asyncpg preventing connection blocking.
• OpenCV pre-processing filters for quick scanned-image OCR extraction.
• State caching with Zustand minimizing redundant API round-trips.

Metrics

Metric	Value
Text Extraction Speed	< 1.2s
AI Resume Generation	2.5s - 4.5s
Editable DOCX Compile	< 150ms
Lighthouse Accessibility	98+

---

♿ Accessibility

• Native HTML5 semantic tags.
• Full keyboard navigation for interactive forms and custom panels.
• Direct compliance with standard high-contrast accessibility color profiles.
• ARIA descriptor attributes added to custom sliders and tabs.

---

📱 Browser Support

Browser	Support
Chrome	✅
Firefox	✅
Safari	✅
Edge	✅

---

🔧 Configuration

.env

GEMINI_API_KEY=your_gemini_api_key_here
LLM_PROVIDER=gemini # gemini | openai
DATABASE_URL=postgresql+asyncpg://postgres:password@localhost:5432/career_os
SECRET_KEY=your_secure_32_character_string
UPLOAD_DIR=uploads
MAX_UPLOAD_SIZE_MB=10
CHROMA_PATH=chroma_db
FRONTEND_URL=http://localhost:5173
ENVIRONMENT=development

Tailwind

In Tailwind CSS v4, dynamic colors and layout preferences are configured directly inside frontend/src/index.css using the @theme directive:

@theme {
  --color-primary: #1a56db;
  --color-secondary: #0f172a;
  --font-sans: "Inter", sans-serif;
}

---

🤝 Contributing

git checkout -b feature/new-feature
git commit -m "Add feature"
git push

---

📝 License

MIT License

---

🙏 Credits

• FastAPI and Uvicorn
• React and Vite
• Google Gemini API
• python-docx
• Tesseract OCR Project

---

📊 Status

• Active development
• Standard multi-source ingestion operational

---

🎯 Roadmap

✅ Completed

• OCR processing for PDF and image uploads.
• Multi-source document merging and de-duplication.
• Real-time custom resume styling and DOCX downloading.
• Detailed job compatibility analyzer.

🚧 In Progress

• Saved generation history for tailored resumes and cover letters.
• Responsive single-page portfolio deployment.

📅 Planned

• Browser extension for direct job description matching from listing pages.
• Interactive interview simulator using extracted candidate history.

---

⭐ Support

If this project helps you, consider giving it a star.