Saiful Islam
Dustin Beaver
As organizations increasingly adopt AI-powered data experiences in production environments, ensuring the accuracy and reliability of AI-generated answers remains a significant challenge. AI agents can interpret user questions, identify relevant data, and generate queries, but limitations in reasoning, context, metadata understanding, and query generation can lead to inaccurate or incomplete results. The Denodo Data Marketplace chatbot addresses this challenge by combining conversational AI with governed enterprise data, metadata, and semantic context. This helps organizations to quickly and efficiently improve the accuracy of AI agents, within an end-user data experience, while maintaining governance and trust. By leveraging the Denodo AI data layer, semantic metadata, and targeted AI-agent improvements, organizations can iteratively enhance chatbot accuracy without requiring major changes to the underlying data infrastructure.
Saiful Islam and Dustin Beaver have come together to demonstrate the latest Denodo capabilities and explore ways that organizations can quickly improve the accuracy of their AI use cases. This session will cover recently released capabilities, including the improved AI context feature included in the Data Marketplace chatbot, called Agent Skills. They will also explain how Agent Skills are able to pull additional context from the metadata of the Data Marketplace, like Property Groups and Custom Elements.
Attend & Learn:
- How to configure an Agent Skill
- How an Agent Skill is used in the Data Marketplace chatbot
- Agent Skill best practices
- Custom Property groups for additional metadata context
