Talk to Your Data embeds Concierge directly inside a Tableau or Power BI Report Viewer, with support for additional platforms on the way. Ask questions in natural language and get answers drawn from the data behind that Report, without leaving the visualization, without writing queries, and without involving an analyst.

Your questions are not limited to what is already visible on a specific visualization. You can explore Dimensions and Filter combinations that the current visualization doesn't show, surfacing analysis that would otherwise require building a new Report or requesting modifications.

Concierge translates natural language into the specific query logic required by the underlying tool (VizQL for Tableau, DAX for Power BI) and executes it directly against the semantic model. No MCP connection or external agentic configuration is required. Answers reflect comprehensive data from the underlying model, respect the access controls already configured in your BI tool, and are grounded in the certified, governed reporting your organization already trusts.

What you can do:

• Ask follow-up questions about any metric, Dimension, or trend in the Report;
• Explore dimensioned data available in the underlying model but not represented in the current visualization;
• Get direct answers without switching tools or waiting for a new Report to be built;
• Trust that answers reflect certified data, not raw or ungoverned sources.

This is a standout feature for Metric Insights, acting as an "Aggregator of Agents." Concierge updates the agent harness providing a unified AI interface across your entire enterprise BI and data ecosystem. Rather than navigating to each platform's own AI tool or knowing in advance which system holds the answer, you ask one question and Concierge determines which agent or agents are best positioned to answer it, dispatches the query, and returns a single consolidated response.

This addresses a fundamental limitation of standalone BI tool agents like Copilot or Tableau Pulse, and database agents like Databricks Genie or Snowflake Cortex: each operates in isolation, with no visibility into the other tools your organization relies on. Conversational analytics platforms like Tellius and ThoughtSpot take a different approach but introduce a different problem. They require you to rebuild your entire semantic layer inside their platform before the AI can function. That context starts incomplete and drifts out of sync the moment anything changes in your actual BI tools. Concierge acts as the connective layer without requiring any of that, working directly against the semantic models and certified reports you already have. No redesign, no migration, no parallel context to maintain.

Connected platforms can include Tableau, Power BI, Databricks, Microsoft Fabric, Snowflake, and others. Where multiple sources are relevant, Concierge queries them in parallel and synthesizes the results into a single response with cited sources. Every answer is informed by your organization's business glossary and certified report catalog, so the context behind the response reflects how your business actually defines and measures performance.

This is the AI equivalent of what Metric Insights already does for BI assets: consolidating content from multiple platforms into one governed portal. Unified AI for BI does the same for AI-assisted analysis.

What you can do:

• Ask one question across all connected data platforms without knowing where the answer actually lives;
• Receive a unified response rather than separate outputs from separate tools;
• Expand coverage automatically as new platforms and agents are added to your environment.

Most BI tools are reactive by design. They wait for a user to open a dashboard, apply filters, and notice something worth investigating. By the time that happens, the opportunity to act has often already passed. Proactive AI Insights changes that relationship between users and their data.

When a significant change is detected in a KPI or metric, Concierge provides AI enhanced detection and alerting. It performs root cause analysis automatically, generates a natural language explanation of what changed and why, and delivers that insight directly to the user through email, Slack, Microsoft Teams, or within the Metric Insights portal. The insight arrives with the relevant dashboard context attached, so users can verify the finding against the certified reports they already trust.

Clicking an Alert does not take you to a raw chart or a data table. It takes you to an AI-generated summary of what changed, why it likely changed, and what the surrounding conditions look like. That explanation is grounded in Concierge's learned context about the metric and domain, and in the governed, certified data behind your existing Reports. Every proactive insight is traceable back to a trusted source, which is what makes it actionable rather than just informational.

Because Concierge monitors the data behind certified dashboards continuously, this capability extends across your entire BI ecosystem. An anomaly in a Tableau Dataset, a Power BI semantic model, or a connected cloud database can all trigger the same governed, explainable alert experience.

What you get:

• Alerts that arrive with explanations, not just notifications;
• Root cause analysis delivered automatically, without requiring manual investigation;
• Insights grounded in certified Reports and your organization's business context;
• Proactive monitoring across all connected platforms, not just a single tool or dashboard;
• Insight available at the moment of awareness, not after further digging.