adithya7reddy/safa repository overview

⁠SAFA - Sentiment Analysis & Feedback Analyzer

A Flask-based web application that analyzes customer reviews using AI-powered sentiment analysis, categorization, and generates actionable insights.

⁠Features

• 📊 Review Summarization: Automatically summarizes customer reviews
• 🏷️ Smart Tagging: Categorizes reviews into relevant themes
• 😊 Sentiment Analysis: Identifies positive, negative, and neutral sentiments
• 📈 Satisfaction Scoring: Calculates customer satisfaction scores (1-100)
• 📝 Analysis Reports: Generates comprehensive business analysis reports
• 💡 Improvement Suggestions: Provides actionable recommendations

⁠Quick Start

⁠Prerequisites

• Docker installed on your system
• Groq API keys (get free keys at https://console.groq.com⁠)

⁠Pull the image

You need to have the image in your local host system to create an instance of the image (container)

docker pull adithya7reddy/safa:2.0

⁠Run the Container

docker run -d \
  -p 8000:5000 \
  -v $(pwd)/uploads:/app/uploads \
  -e summarize_key="your_groq_api_key" \
  -e tag_key="your_groq_api_key" \
  -e analysis_key="your_groq_api_key" \
  -e improvements_key="your_groq_api_key" \
  --name safa_container \
  adithya7reddy/safa:2.0

⁠Access the Application

Open your browser and navigate to: http://localhost:8000⁠

⁠Environment Variables

The application requires 4 API keys and 4 model names (with defaults):

⁠Using Different Models

You can specify different AI models for each function:

docker run -d \
  -p 8000:5000 \
  -v $(pwd)/uploads:/app/uploads \
  -e summarize_key="gsk_xxxxx" \
  -e summarize_model="llama-3.3-70b-versatile" \
  -e tag_key="gsk_xxxxx" \
  -e tag_model="mixtral-8x7b-32768" \
  -e analysis_key="gsk_xxxxx" \
  -e analysis_model="qwen/qwen3-32b" \
  -e improvements_key="gsk_xxxxx" \
  -e improvements_model="gemma2-9b-it" \
  --name safa_container \
  adithya7reddy/safa:2.0

⁠Using Environment File

For easier management, create a .env file:

summarize_key=your_api_key1_here
summarize_model=llama-3.3-70b-versatile
tag_key=your_api_key2_here
tag_model=llama-3.3-70b-versatile
analysis_key=your_api_key3_here
analysis_model=qwen/qwen3-32b
improvements_key=your_api_key4_here
improvements_model=qwen/qwen3-32b

Then run:

docker run -d
-p 8000:5000
-v $(pwd)/uploads:/app/uploads
--env-file .env
--name safa_container
adithya7reddy/safa:2.0

⁠Input Format

Upload an Excel file (.xlsx) with the following columns:

• Text: Customer review text
• Rating: Numerical rating (optional)

Example:

⁠Supported Models

Compatible Groq models include:

• llama-3.3-70b-versatile
• llama-3.1-70b-versatile
• mixtral-8x7b-32768
• gemma2-9b-it
• qwen/qwen3-32b

Check Groq documentation⁠ for the latest available models.

⁠Ports

• Container Port: 5000
• Host Port: 8000 (configurable with -p flag)

⁠Volume Mounting

The -v $(pwd)/uploads:/app/uploads flag ensures uploaded files persist between container restarts.

⁠Container Management

Stop the container:

docker stop safa_container

Start the container:

docker start safa_container

View logs:

docker logs safa_container

Remove the container:

docker rm safa_container

⁠Troubleshooting

Problem: Application won't start

• Solution: Check if all required API keys are provided and all the requirements are met
• Run docker logs safa_container to see error messages

Problem: Cannot access localhost:8000

• Solution: Ensure port 8000 is not in use by another application
• Change the port: -p 9000:5000

Problem: Pydantic errors

• Solution: This version includes Pydantic 2.9.2 for compatibility

⁠Technology Stack

• Backend: Flask 3.1.0
• AI/ML: LangChain, Groq API
• Data Processing: Pandas, NumPy
• Containerization: Docker

⁠Version History

• 2.0: Environment variable support for API keys and models
• 1.0: Initial release for testing - Recommended not to use

⁠License

This project is open source and available for educational purposes.

⁠Support

For issues or questions, please open an issue on the GitHub repository: Adithya2369/safa_public