> ## Documentation Index > Fetch the complete documentation index at: https://docs.trymaitai.com/llms.txt > Use this file to discover all available pages before exploring further. # Visualizer > Mapping out potential conversation paths and logic Generating high-quality training data is often the biggest bottleneck in building great AI. **Synthetic Conversation Trees (SCTs)** solve this by allowing you to map out possible interactions visually. SCT Visualizer ### What is a Conversation Tree? A **Conversation Tree** is a visual map of all the potential paths a conversation can take. It consists of: * **Nodes**: Specific states in the conversation (e.g., "User asks for refund"). * **Edges**: Transitions between states (e.g., "Agent asks for order number"). * **Variables**: Placeholders for dynamic data (e.g., order\_id and customer\_name wrapped in double curly braces) that ensure diversity in the generated output. ### Why map it visually? The visualizer allows you to ensure logic completeness. By mapping out the tree, you can easily spot missing edge cases or circular logic that would otherwise be difficult to identify in raw text datasets. ## Next * Generate synthetic data from the tree: [Data Generation](/build/scts/data_generation)