Chat directly with your AI agents using natural language to get targeted answers to specific business questions.

This new chat interface enables data analysts to quickly investigate specific questions without navigating through multiple UI elements. Instead of clicking through tabs and filters to explore data correlations, you can now ask your agent direct questions like:

The chat maintains conversation context, allowing you to ask follow-up questions and build deeper investigations naturally. Each agent includes pre-defined question suggestions to help you get started.

Follow your existing discovery workflow - use the traditional UI to browse highlights and discover patterns, then switch to chat when you need to investigate specific questions or prepare answers for stakeholder meetings.

Want to get started? Open any agent and look for the "Chat with agent" button to begin asking questions about your data.

🧪 We're actively expanding query capabilities and would love your feedback on what works well and what could be better.

We now support two distinct analysis modes, giving you the flexibility to analyze your data exactly how your business needs it.

Compare different time periods to see what's changed between them, or analyze a single timeframe to identify emerging patterns and trends within that period.

The Highlights report automatically adapts to show the most relevant insights based on your selected approach.

Access the new date analysis options in the top bar of any agent dashboard.

We've redesigned the Highlights Report to transform raw insights into executive-ready business stories with clear next steps.

The new version goes beyond showing what changed to explaining why it matters and what you should do about it. Instead of basic trend data, you now get compelling narratives like "Privacy concerns became your #5 churn driver with new subscribers showing 289% higher cancellation rates among enterprise healthcare customers" followed by specific AI-powered recommendations.

Perfect for data analysts who need to confidently present insights to leadership teams. The enhanced report turns complex Snowflake data into actionable business intelligence that drives immediate decision-making.

Each insight now tells a complete story: what changed, which customers are affected, and exactly what your team should do about it.

Find the updated "Highlights Report" in your agent dashboard to start creating compelling data stories.

Get comprehensive understanding of customer issues without reading through raw feedback. The new Insight Summary automatically identifies trends, root causes, and representative quotes.

The summary breaks down topic evolution over time, affected customer segments, and provides actual customer quotes to add context. This helps you quickly understand the full picture behind each insight without manual analysis.

Access Insight Summary by clicking on any insight in your dashboard to see the complete analysis.

Quickly identify emerging issues and trends across your Snowflake data. The new Highlights Report automatically surfaces the most significant changes.

The report shows your highest increasing and decreasing topics, plus changes in metadata distributions like product categories and brands. This helps you spot emerging problems before they escalate and identify what's resonating with customers.

Look for the "Highlights Report" in your dashboard to start tracking customer feedback trends.

Your preferred date ranges now persist between sessions, with preset options like "Last 7 days" available for faster reporting. 

Date filters automatically apply when you reopen the app.

Launched flexible date range selection with period-over-period comparison, allowing teams to analyze trends across any custom time-frame instead of fixed periods.

Added ranking movement indicators (↑↓) to show how customer interaction categories change in importance over time, helping teams quickly spot and communicate emerging patterns.

We're upgrading our AI engine to Claude 3.5 for enhanced accuracy and reliability in your AI agent insights. This brings significant improvements to our core analysis capabilities.

Claude 3.5 is now available for customers using Snowflake in the Oregon region. For other regions, you can enable Cross-region inference in your Snowflake account to access Claude 3.5 (see docs).

We've added timezone configuration to help global teams work with their data more accurately. Your AI agents will now correctly interpret dates across different timezones.

The new timezone settings include automatic detection from your data and manual override options when needed. This eliminates date discrepancies when analyzing data across different regions.

Configure your timezone in your agent's settings.