Spaces:

CareerAI
/

app

Running

App Files Files Community

app / README.md

CareerAI-app

Deploy CareerAI to HuggingFace Spaces

b7934cd 11 days ago

preview code

raw

history blame contribute delete

15.6 kB

metadata

title: CareerAI
emoji: 🚀
colorFrom: green
colorTo: blue
sdk: docker
app_port: 7860
pinned: true

CareerAI

🧠 AI-Powered Career Assistant | Asistente Inteligente de Carrera
Analyze your CV · Generate Cover Letters · Simulate Interviews · Search Jobs

🇺🇸 English · 🇦🇷 Español

🇺🇸 English

🎬 What is CareerAI?

CareerAI is an AI-powered web application that helps you boost your professional career. Upload your documents (CV, cover letters, certificates) and the AI assistant analyzes them using advanced Retrieval-Augmented Generation (RAG) to give you personalized recommendations, generate professional documents, and prepare you for job interviews.

✨ 100% Free · No hallucinations · Based on your real documents

🚀 Key Features

🤖 Custom AI Models

Model	Engine	Description
🧠 CareerAI Pro	Llama 3.3 70B	Maximum quality · Detailed responses
⚡ CareerAI Flash	Llama 3.1 8B	Ultra fast · Instant responses

💬 5 Assistant Modes

Mode	What it does
💬 General Chat	Ask anything about your professional career
🎯 Job Match	Analyze your compatibility with job offers (% match)
✉️ Cover Letter	Generate personalized cover letters using your real CV
📈 Skills Gap	Identify missing skills + roadmap to improve
🎤 Interview	Simulate interviews with technical and STAR method questions

📋 Full Feature List

Feature	Description
📄 Multi-format	Supports PDF, DOCX, TXT, images (JPG, PNG, WebP)
🖼️ Vision AI	Smart reading of scanned PDFs and document photos
⚡ Streaming	Real-time token-by-token responses
📤 Premium Export	Export to PDF, DOCX, HTML, TXT with professional formatting
📊 Dashboard	Skills charts, professional timeline, AI insights
🔐 Full Auth	Register, login, Google OAuth, password reset
💼 Job Search	Integration with LinkedIn, Indeed, Glassdoor via JSearch
🎨 Premium UI	Claude/ChatGPT-style design with dark mode
📱 Responsive	Works on desktop, tablet, and mobile
💾 Persistence	Chat history synced to the cloud

🧠 RAG Pipeline v2.0

CareerAI uses an advanced retrieval pipeline combining multiple techniques to find the most relevant information from your documents:

📝 User Query
     │
     ├── 1️⃣ Vector Search (Semantic)
     │      └── ChromaDB + BGE-M3 (100+ languages)
     │
     ├── 2️⃣ Keyword Search (Lexical)
     │      └── BM25 lexical matching
     │
     ├── 3️⃣ Reciprocal Rank Fusion (RRF)
     │      └── Merges semantic + lexical results
     │
     ├── 4️⃣ Reranking (Cross-Encoder)
     │      └── BGE-Reranker-v2-m3 (relevance reordering)
     │
     └── 5️⃣ LLM with optimized context
            └── Groq + Llama 3.3 70B (streaming)

Available Embedding Models

Model	Languages	Size	Performance
🌍 BGE-M3 (Recommended)	100+	~2.3 GB	⭐⭐⭐⭐⭐
🚀 GTE Multilingual	70+	~580 MB	⭐⭐⭐⭐
📐 Multilingual E5	100+	~1.1 GB	⭐⭐⭐⭐
⚡ MiniLM v2	English	~90 MB	⭐⭐⭐

🛠️ Tech Stack

Layer	Technology
Backend	FastAPI + Uvicorn
Frontend	HTML5 + CSS3 + JavaScript (Claude-style)
LLM	Groq API (Llama 3.3 70B / Llama 3.1 8B)
RAG	ChromaDB + BM25 + BGE-M3 + Reranker + RRF
Database	SQLite + SQLAlchemy
Auth	JWT + BCrypt + Google OAuth
Email	FastAPI-Mail + Gmail SMTP
Vision AI	Groq + Llama 4 Scout
Embeddings	HuggingFace (BGE-M3, GTE, E5, MiniLM)
Export	FPDF2, python-docx
Job Search	JSearch API (RapidAPI)

� Installation & Setup

1. Clone the repository

git clone https://github.com/Nicola671/CareerAI.git
cd CareerAI

2. Create virtual environment

python -m venv venv

# Windows
venv\Scripts\activate

# Mac/Linux
source venv/bin/activate

3. Install dependencies

pip install -r requirements.txt

4. Configure environment variables

Create a .env file in the project root:

# Groq API Key (free from console.groq.com)
GROQ_API_KEY=your_api_key_here

# JWT Secret (change to something random)
SECRET_KEY=your_very_long_random_secret_key

# Email for password recovery (optional)
MAIL_USERNAME=your_email@gmail.com
MAIL_PASSWORD=your_app_password
MAIL_FROM=your_email@gmail.com

# JSearch API Key for job search (optional)
JSEARCH_API_KEY=your_jsearch_key

5. Get Groq API Key (FREE)

Go to console.groq.com
Create a free account
Go to "API Keys" → "Create API Key"
Copy your key (starts with gsk_...)
Paste it in your .env file

6. Run

uvicorn api:app --reload --port 8000

Open http://localhost:8000 in your browser 🚀

� API Endpoints (22 routes)

Group	Endpoints	Description
🏠 Frontend	`GET /`	Serves the web app
⚙️ Config	`GET /api/status`, `POST /api/config`	Status & configuration
💬 Chat	`POST /api/chat`, `POST /api/chat/stream`	Chat with/without streaming
📄 Docs	`POST /api/documents`, `GET /api/documents`, `DELETE /api/documents/{file}`	Document CRUD
📤 Export	`POST /api/export`, `POST /api/export/conversation`	Export to PDF/DOCX/HTML/TXT
💼 Jobs	`GET /api/jobs`	Job search
📊 Dashboard	`GET /api/dashboard`	AI-powered profile analysis
🔐 Auth	`POST /api/auth/register`, `POST /api/auth/login`, `GET /api/auth/me`	Full authentication

Interactive API docs: http://localhost:8000/docs (Swagger UI)

📊 Project Metrics

Metric	Value
Lines of code	8,400+
API Endpoints	22
Frontend functions	80+
Backend functions	60+
Assistant modes	5
Export formats	4 (PDF, DOCX, HTML, TXT)
Upload formats	7 (PDF, DOCX, TXT, JPG, PNG, WEBP)
Embedding models	4

🇦🇷 Español

🎬 ¿Qué es CareerAI?

CareerAI es una aplicación web de inteligencia artificial que te ayuda a impulsar tu carrera profesional. Subís tus documentos (CV, cartas, certificados) y el asistente los analiza con IA avanzada (RAG) para darte recomendaciones personalizadas, generar documentos profesionales y prepararte para entrevistas.

✨ Todo esto 100% gratis · Sin alucinaciones · Basado en tus documentos reales

🚀 Funcionalidades Principales

🤖 Modelos de IA Personalizados

Modelo	Motor	Descripción
🧠 CareerAI Pro	Llama 3.3 70B	Máxima calidad · Respuestas detalladas
⚡ CareerAI Flash	Llama 3.1 8B	Ultra rápido · Respuestas al instante

💬 5 Modos del Asistente

Modo	Qué hace
💬 Chat General	Consultá lo que quieras sobre tu carrera profesional
🎯 Job Match	Analizá tu compatibilidad con ofertas de trabajo (% de match)
✉️ Cover Letter	Generá cartas de presentación personalizadas usando tu CV real
📈 Skills Gap	Identificá habilidades faltantes + roadmap para mejorar
🎤 Entrevista	Simulá entrevistas con preguntas técnicas y método STAR

📋 Lista Completa de Características

Feature	Descripción
📄 Multi-formato	Soporta PDF, DOCX, TXT, imágenes (JPG, PNG, WebP)
🖼️ Vision AI	Lectura inteligente de PDFs escaneados y fotos de documentos
⚡ Streaming	Respuestas en tiempo real token por token
📤 Export Premium	Exportá a PDF, DOCX, HTML, TXT con formato profesional
📊 Dashboard	Gráficos de skills, timeline profesional, insights de IA
🔐 Auth Completo	Registro, login, Google OAuth, reset de contraseña
💼 Búsqueda de Empleo	Integración con LinkedIn, Indeed, Glassdoor via JSearch
🎨 UI Premium	Diseño tipo Claude/ChatGPT con dark mode
📱 Responsive	Funciona en desktop, tablet y celular
💾 Persistencia	Historial de chats sincronizado en la nube

🧠 Pipeline RAG v2.0

CareerAI usa un pipeline de retrieval avanzado que combina múltiples técnicas para encontrar la información más relevante de tus documentos:

📝 Query del usuario
     │
     ├── 1️⃣ Vector Search (Semántico)
     │      └── ChromaDB + BGE-M3 (100+ idiomas)
     │
     ├── 2️⃣ Keyword Search (Léxico)
     │      └── BM25 lexical matching
     │
     ├── 3️⃣ Reciprocal Rank Fusion (RRF)
     │      └── Combina resultados semánticos + léxicos
     │
     ├── 4️⃣ Reranking (Cross-Encoder)
     │      └── BGE-Reranker-v2-m3 (reordena por relevancia)
     │
     └── 5️⃣ LLM con contexto optimizado
            └── Groq + Llama 3.3 70B (streaming)

🛠️ Stack Tecnológico

Capa	Tecnología
Backend	FastAPI + Uvicorn
Frontend	HTML5 + CSS3 + JavaScript (estilo Claude)
LLM	Groq API (Llama 3.3 70B / Llama 3.1 8B)
RAG	ChromaDB + BM25 + BGE-M3 + Reranker + RRF
Base de datos	SQLite + SQLAlchemy
Auth	JWT + BCrypt + Google OAuth
Email	FastAPI-Mail + Gmail SMTP
Vision AI	Groq + Llama 4 Scout
Embeddings	HuggingFace (BGE-M3, GTE, E5, MiniLM)
Exportación	FPDF2, python-docx
Búsqueda	JSearch API (RapidAPI)

🚀 Instalación y Setup

1. Clonar el repositorio

git clone https://github.com/Nicola671/CareerAI.git
cd CareerAI

2. Crear entorno virtual

python -m venv venv

# Windows
venv\Scripts\activate

# Mac/Linux
source venv/bin/activate

3. Instalar dependencias

pip install -r requirements.txt

4. Configurar variables de entorno

Creá un archivo .env en la raíz del proyecto:

# Groq API Key (gratis desde console.groq.com)
GROQ_API_KEY=tu_api_key_aqui

# JWT Secret (cambiá por algo aleatorio)
SECRET_KEY=tu_secret_key_muy_larga_y_aleatoria

# Email para recuperación de contraseña (opcional)
MAIL_USERNAME=tu_email@gmail.com
MAIL_PASSWORD=tu_app_password
MAIL_FROM=tu_email@gmail.com

# JSearch API Key para búsqueda de empleos (opcional)
JSEARCH_API_KEY=tu_jsearch_key

5. Obtener API Key de Groq (GRATIS)

Andá a console.groq.com
Creá una cuenta gratis
Andá a "API Keys" → "Create API Key"
Copiá tu key (empieza con gsk_...)
Pegala en el archivo .env

6. Ejecutar

uvicorn api:app --reload --port 8000

Abrí http://localhost:8000 en tu navegador 🚀

� Estructura del Proyecto

CareerAI/
├── api.py                        # 🚀 Backend FastAPI (22 endpoints)
├── requirements.txt              # 📦 Dependencias Python
├── .env                          # 🔐 Variables de entorno (NO se sube a Git)
├── README.md                     # 📖 Este archivo
│
├── frontend/                     # 🎨 UI tipo Claude
│   ├── index.html                # Estructura HTML
│   ├── app.js                    # Lógica completa (1,842 líneas)
│   ├── styles.css                # Sistema de diseño (1,695 líneas)
│   ├── icon-pro.png              # 🧠 Icono CareerAI Pro
│   ├── icon-flash.png            # ⚡ Icono CareerAI Flash
│   └── favicon.png               # Favicon
│
├── src/                          # 🧠 Core Engine
│   ├── career_assistant.py       # Motor IA con 5 modos especializados
│   ├── rag_engine.py             # RAG v2.0 (Hybrid + Reranking + RRF)
│   ├── document_processor.py     # Procesador multi-formato + Vision AI
│   ├── profile_extractor.py      # Extractor de perfil para dashboard
│   ├── exporter.py               # Exportación PDF/DOCX/HTML/TXT
│   ├── auth.py                   # Autenticación (JWT + Google OAuth)
│   └── models.py                 # Modelos SQLAlchemy (User, Conversation)
│
└── data/                         # 💾 Datos (no se suben a Git)
    ├── uploads/                  # Documentos subidos
    └── vectordb/                 # ChromaDB persistencia

🆓 ¿Por qué es 100% Gratis? / Why is it 100% Free?

Component	Cost
Groq API (Llama 3.3 70B)	✅ Free (generous rate limits)
BGE-M3 Embeddings	✅ Free (runs locally)
BGE-Reranker-v2-m3	✅ Free (runs locally)
BM25 Keyword Search	✅ Free (runs locally)
ChromaDB Vector Store	✅ Free (runs locally)
FastAPI + Frontend	✅ Free (open source)
SQLite Database	✅ Free (runs locally)

🤝 Contributing

Contributions are welcome! If you want to improve CareerAI:

Fork the repository
Create a branch: git checkout -b feature/new-feature
Commit: git commit -m "Add new feature"
Push: git push origin feature/new-feature
Open a Pull Request

📄 License

This project is licensed under the MIT License. Feel free to use, modify, and distribute it.

👨‍💻 Author / Autor

Nicolás Medina

If this project helped you, consider giving it a ⭐ on GitHub!
Si este proyecto te ayudó, ¡considerá darle una ⭐ en GitHub!

CareerAI v1.0 — FastAPI + RAG v2.0 + Groq
Made with ❤️ in Argentina 🇦🇷