Context Suggestions

Overview

The AI Context Suggestions feature analyzes your chatbot's failed conversations to identify patterns and provide actionable recommendations for improvement. By examining unanswered questions and failed responses, the system generates targeted suggestions to enhance your chatbot's knowledge base and performance.

Regularly generating suggestions helps you maintain an effective chatbot by addressing the most common knowledge gaps and user issues.

How It Works

The system analyzes recent failed chatbot interactions to:

Identify patterns - Find recurring issues and unanswered questions
Prioritize improvements - Focus on the most frequent problems first
Generate recommendations - Provide specific suggestions for your knowledge base
Suggest context tools - Recommend API integrations when dynamic data is needed

Suggestions are generated based on the last 500 failed responses since your previous analysis, ensuring you get fresh insights as your chatbot evolves.

Generating Suggestions

To generate new suggestions for your chatbot:

Navigate to Suggestions

Go to your chatbot's Automation > AI > Suggestions page in the dashboard.

Click Generate

Click the "Generate suggestions" button to start analysis.

Wait for Analysis

The system will:

Analyze recent failed responses
Process patterns using AI
Stream results in real-time

Review Results

View the generated suggestions, which include:

Summary of main patterns found
Prioritized knowledge gaps to address
Context tool recommendations (when applicable)

Analysis requires at least 10 new failed responses since your last suggestions. If you have fewer events, you'll need to wait for more conversation data.

Viewing Previous Suggestions

All previously generated suggestions are displayed chronologically, showing:

Generation timestamp - When the analysis was performed
Analysis summary - Overview of patterns identified
Knowledge gaps - Specific questions or topics to add to your knowledge base
Context tool ideas - API integration suggestions for dynamic data needs

Best Practices

Generate regularly - Run suggestions analysis weekly or after significant chatbot updates
Review patterns - Focus on the most frequent issues first for maximum impact
Update knowledge base - Add the suggested questions and answers to improve response accuracy
Consider API tools - Implement context tools when suggestions identify dynamic data needs

The AI prioritizes quality over quantity, focusing on the most impactful improvements that will resolve the largest number of failed conversations.