Conversational Analytics with Kyvos Dialogs

This section describes the capabilities of Kyvos Dialogs powered by AI and explains how it enables conversational analytics on governed enterprise data. It outlines key features, workflow, and the available offerings to help users understand how Kyvos Dialogs can be used in both evaluation and production environments. 

Kyvos Dialogs Architecture

Before starting with Kyvos Dialogs

To use Kyvos Dialogs with all features,

• You need a valid license file. Please refer to your license agreement or contact your Kyvos representative for specific details about your plan or rates. Once you obtain the license from your Kyvos representative, you must upload it to the Kyvos License page through Kyvos Manager.

• You need to create a AI connection and embedding connection. To navigate to the Kyvos connection page and create a AI LLM connection.

• You need to complete the prerequisites for:

  • Web-based: Refer to the prerequisites section.

  • Desktop-based: Refer to the prerequisites section.

LLM Connection, Embedding Connection, and Models

• LLM Connection is the configuration that enables a system to communicate with a Large Language Model (LLM), such as those developed by OpenAI, Azure OpenAI, AWS Bedrock, and Vertex AI, so on.

• An embedding connection in Kyvos refers to the configuration that links the platform to an embedding model—a machine learning model that transforms textual or complex data into dense, numerical vectors called embeddings.

• Certified LLM and Embedding Models

Source of Conversational Analytics with Kyvos Dialogs

How Kyvos Dialogs Works

Kyvos Dialogs allows users to ask questions in natural language and receive accurate, contextuais insights—without needing technical expertise.

1. Ask a Question: Users submit questions in plain language, with no need to understand schemas or write queries.

2. Understand the Intent: Kyvos uses its semantic layer and Retrieval-Augmented Generation (RAG) to accurately interpret business terms and data relationships.

3. Generate and Execute the Query: A large language model converts the request into an optimized query, which Kyvos executes at scale across enterprise data sources.

4. Deliver Insights: Results are summarized and presented automatically as interactive visualizations or tables, with options for customization.

Kyvos Dialogs delivers fast, reliable insights through a natural language interface, making data analysis simple, intuitive, and efficient.

Let’s take a quick look at Kyvos Dialogs.

Open

Let's start with a quick tour of your Kyvos Dialogs

Open

Features of Kyvos Dialogs

• Natural Language Querying: Users can ask questions using plain business language without needing knowledge of SQL, coding, or underlying data structures. This simplifies data access, enables independent analysis of large datasets, and reduces reliance on technical teams for routine queries.

• Contextual exploration: Kyvos Dialogs preserves conversational context across queries within a session. Users can ask follow-up questions, refine previous queries, or explore related dimensions without restarting the analysis.

• AI-Powered Visualizations: Get answers to all your queries in the form of best-fit visualizations along with a natural language summary revealing key insights.

• Real-time KPI generation: Users can create, modify, and refine business metrics dynamically during a conversation. KPIs are updated in real time based on current data, allowing users to focus on the metrics most relevant to their decisions.

• Guided data discovery: Kyvos Dialogs automatically identifies the most relevant semantic models and recommends appropriate visualizations. By considering data types, hierarchies, and user intent, it presents insights in formats that are easy to understand and act upon.

• Reuse Chat Sessions Across Organization: Share chat sessions across users and groups within your organization to reuse analytics summaries for quicker decision-making.

• Data Governance: Use the power of AI while protecting your data with Kyvos' multi-layered security architecture, ensuring strong privacy protection.

• Automated Processes: Use AI-driven analytics to streamline workflows, significantly reducing the time and effort required for in-depth data discovery.

• Text-to-SQL: It converts user input questions into the appropriate SQL query language. 

Next: Select your Kyvos Dialogs Platform: Web-based or Desktop-based

Read also: