OVERVIEW

ODC Chunking Library Demo is a sample ODC application that shows how to use the ODC Chunking Library in a real app. It lets you load sample Markdown or plain text, adjust chunking parameters, run different chunking strategies, and inspect the returned results. The app is intended as a practical reference for developers who want to understand how to connect the library to screens, actions, user inputs, and result visualization before integrating chunking into their own applications.

PURPOSE

This demo application is designed to help developers:

• Understand how to invoke the chunking library from an ODC app
• Compare different chunking strategies using the same source content
• Experiment with parameters such as chunk size, overlap, and structure-preservation options
• Review the returned chunk data, metadata, and summary statistics
• Use the sample as a starting point for their own AI, search, or retrieval scenarios

WHAT THE DEMO INCLUDES

• A screen to run chunking interactively
• Sample Markdown and plain text content
• Support for multiple chunking strategies:
  • Markdown
  • Recursive
  • Sentence
  • Character
• Editable chunking parameters
• Result visualization with chunk text and metadata
• JSON download of the generated output
• Example server-side orchestration for strategy selection and validation

HOW THE DEMO WORKS

The application loads either sample Markdown or sample text content and lets the user choose a chunking strategy. After the user sets the desired parameters and runs the process, the app calls the chunking logic and displays the returned chunks on screen. Each result includes the generated chunk text, metadata such as character range and token estimate, and summary information about the full operation. The app also allows downloading the generated result as JSON.

MAIN SCREEN

The demo centers around a single interactive screen where developers can:

• Select the chunking strategy
• Choose whether to use Markdown or plain text sample content
• Adjust settings such as:
  • chunk size
  • overlap size
  • max total characters
  • sentence limits
  • whitespace normalization
  • Markdown preservation options
• Run chunking and inspect the output
• Download the result as JSON

WHAT DEVELOPERS CAN LEARN FROM THIS DEMO

• How to structure screen inputs for chunking scenarios
• How to pass parameters from the UI into server logic
• How to validate chunking inputs before execution
• How to route execution to different chunking strategies
• How to display chunk collections and metadata in the UI
• How to expose chunking output for export or downstream processing

DEMO USE CASES

This sample is useful for developers building:

• AI-ready document preprocessing flows
• Semantic search preparation workflows
• Retrieval-augmented application features
• Knowledge base indexing pipelines
• Prototypes that need chunk inspection before embedding or indexing

LIMITATIONS

• This is a demonstration app focused on showing integration patterns
• It uses sample content rather than a full production ingestion flow
• It does not generate embeddings or store vectors
• It does not include indexing or vector database integration
• Token values shown in results are estimates
• Final parameter tuning depends on the document type and downstream usage

WHEN TO USE THIS DEMO

Use this application when you want a working example of how to integrate chunking into an ODC app. It is especially useful for learning, testing, validating parameter choices, and accelerating the implementation of your own chunk-based workflows.

NEXT STEPS

After exploring the demo, developers can adapt the same patterns to:

• Replace sample text with real uploaded or stored documents
• Send returned chunks into embedding pipelines
• Store chunk output in their own data model
• Integrate with search or retrieval services
• Extend the UI for production scenarios