Guidance Throughout & Even Simpler Chat Experience
Making products simpler is a harder task than adding features. Our Valentine’s Day release is all about giving customers what they asked for, and this time the overall theme is guidance and simplicity. With this release, we’ve eliminated several steps to streamline the conversational experience for advanced analytics. Previously you had to set up the analysis by first choosing the variable you wanted to understand before you could chat. Now, you can jump right into your conversation.
Automatic KPI Detection & Problematic Predictor Guidance
Misleading factors can hamper analytics when exploring multiple variables, and many users also need deeper KPI insights without realizing it. Our new release addresses both. Automatic Problematic Predictor Guidance identifies predictors that may distort results and suggests which factors to exclude, boosting analytical accuracy and preventing costly misinterpretations. Simultaneously, Automatic KPI Detection recognizes when your conversation implies a KPI-focused query and seamlessly transitions into an augmented analytics chat, removing the guesswork of manual switching.
Auto-Recommend Context (Data) Search in Chat
We know that half the battle is often finding the right data. To tackle this, we’ve improved the Auto-Recommend Context Dataset Search. After submitting a question, our system automatically scans for relevant keywords and runs a dataset search in the background. You’ll then see a ranked list of matching data or document sets, along with explanations of why they’re a good fit. This means less time digging around for the right data source and more time uncovering valuable insights. It’s a frictionless way to ensure you’re always tapping into the most relevant information for your analytics needs.
Beta: Adding some of the core improvements of the DeepSeek models, without using such models themselves.
