Conversational Analytics with Kyvos Dialogs

• AI-Powered Visualizations: Automatically selects and generates appropriate charts, tables, and summaries based on the user’s intent, data characteristics, and semantic model.

• Real-Time KPI Generation: Allows users to create, modify, and refine business metrics dynamically during a conversation, enabling more flexible and interactive analysis.

• Guided Data Discovery: Uses the Kyvos semantic layer to identify relevant datasets, dimensions, measures, hierarchies, and relationships, helping users discover and explore enterprise data more effectively.

• Automated Query Generation and Optimization: Translates natural language questions into optimized analytical queries while leveraging the Kyvos semantic layer to ensure that business definitions, calculations, hierarchies, and governed access are correctly applied.

• Deep Dive Analysis: Enables multi-hop analytical reasoning for complex business questions that require multiple steps of investigation. Instead of simply generating a single query and returning an answer, Dialogs can plan and execute a sequence of analytical steps, using tools such as enterprise knowledge bases, web search, and code execution where appropriate. For example, when asked “Why did profit decline last quarter and what should we do?”, Dialogs can investigate the underlying data, analyze contributing factors across products, regions, or other dimensions, retrieve relevant business context, perform additional calculations, and synthesize the findings into a grounded explanation and recommendation.

• Personalized Analytics: Dialogs can learn from user feedback and preferences by incorporating them into a personalized context for future interactions. This allows the analytics experience to become more relevant over time—for example, remembering preferred analytical approaches, terminology, visualization preferences, or areas of investigation.
  Personalization helps move conversational analytics from a generic question-answering experience toward an individualized analytics assistant.

• Enhanced Monitoring, Auditing & Query Analysis: Dialogs provides enhanced observability into how an analytical response was produced. The Natural Language Query Analyzer gives administrators and analysts visibility into the execution of a conversational request, including:

  • Token consumption

  • Time spent across individual processing steps

  • Query generation and execution time

  • Tool calls and their execution time

  • Contribution of different stages to overall response latency

  • Detailed analysis of the conversational workflow

This provides organizations with greater transparency, performance optimization, cost management, troubleshooting, and auditing capabilities for AI-powered analytics.

• Enterprise Data Governance and Security: Dialogs operates within Kyvos’ governed analytics framework, respecting enterprise security policies and role-based access controls. Users receive insights only from data they are authorized to access, helping organizations bring conversational AI to enterprise data without compromising governance.

Jump to the section that matches what you're trying to do:

Kyvos Dialogs provides two analysis modes that allow you to choose the level of reasoning used to answer your queries. Depending on the complexity of your request, you can select either Quick Analysis for fast responses or Deep Dive Analysis for comprehensive, multi-step reasoning.

Kyvos Dialogs is available in two editions: Enterprise and Free.

• Kyvos Dialogs Enterprise: Designed for production deployment connected to existing Kyvos environment. It provides full access to enterprise semantic models, governance controls, scalability, and advanced analytics capabilities.

• Kyvos Dialogs Free: Available through the Kyvos Trial environment, this edition connects to a sample semantic model and allows users to experience conversational analytics without requiring a full Kyvos deployment.

Read also: