Improved RAG Q&A

We made significant improvements to our RAG document processing pipeline that will enhance the quality of responses you receive from our platform.

NOTE: These updates will require a one-time reindexing of your documents to take full advantage of these enhancements.

Improved Document Parsing
Enhanced parsing capabilities with full support for markdown formatting, preserving more of your document's original structure and meaning
Advanced Chunking Strategies
New semantic chunking that intelligently breaks documents based on content meaning rather than character count, plus expanded chunk options for better context retention
Next-Generation Model
Implementation of a much larger generation model delivering more accurate, nuanced, and contextually appropriate responses
Hybrid Search
Enhanced retrieval based on semantic and lexical searches
Performance Improvements
Optimized index organization for improved performance

Improved quality of responses with RAG AI agents.

To avoid service disruption, republish the documents and perform the following steps:

Navigate to My Client Groups > Configure HCM Agents > Tools > Open the AI Agent tool
Check the status of your agent documents: If any are not marked as “Ready to Publish”, update them to “Ready to Publish”
Run the “Process Agent Documents” job from Tools > Scheduled Processes

This will restore the agent functionality.

For more detailed information on how to setup AI Agents, please review