FabriXAI User Guide

# Customer Review Classifier

The Customer Review Classifier is an AI assistant designed to analyze user feedback and categorize it into predefined categories, offering insights into the sentiment (positive, negative, or neutral) associated with each category. It allows businesses to understand customer experiences across various facets, such as product features, user experience, performance, customer support, and more. By automating the analysis of feedback, the classifier aids in identifying improvements and highlighting areas of strength, ultimately enhancing customer satisfaction and retention.

# Getting Started

To get started, make sure you have cloned the Customer Review Classifier. Follow Create from Template to clone and configure your AI agent.

# Configuration Instructions

# Knowledge

You can upload .pdf or .txt files to provide the agent with tailored context that meets your specific requirements.

# User Inputs

Field Title Description Instruction
Feedback Text The text of user feedback to be analyzed. Input should be clear and concise, typically 1-5 sentences.

# AI Models

The default model is Llama 3.1 8B Instruct, but you can explore other supported AI models as well. If needed, refine your prompts for more tailored outputs.

TIP

For more detailed instructions on configuring an AI agent, please refer to Configure AI Agent

# Sample Inputs and Output

# Sample Inputs

  • Feedback Text: The app is very user-friendly, but the loading time can be improved for a better experience.

# Sample Output

{
  "Product Features and Functionality": {
    "Categories": ["Core Features"],
    "Sentiment": "Positive"
  },
  "User Experience and Design": {
    "Categories": ["Ease of Use"],
    "Sentiment": "Positive"
  },
  "Performance and Reliability": {
    "Categories": ["Speed and Responsiveness"],
    "Sentiment": "Negative"
  }
}

# Tips for Effective Configuration

  • Be Specific: Input clear and descriptive feedback to enable accurate categorization and sentiment analysis.
  • Use Natural Language: Write feedback as if you were communicating with another person to allow the AI to interpret it effectively.
  • Limit Input Length: Keep feedback concise to enhance analysis accuracy, ideally between 1-5 sentences.
  • Provide Context: If necessary, upload documents that expand on your specific use case to improve the classifier's understanding.
  • Review Output Regularly: Analyze the categorization and sentiment regularly to refine context and input for better results.

Continue exploring our collection of Agent Templates to discover more AI agent ideas in action!