AI Model Wiki

    Skip to main content
    14 min
    Lesson 2 of 4

    Building AI-Enhanced Dashboards

    From the course Tableau AI & Pulse

    Combine Tableau's traditional visualization power with AI features to create dashboards that don't just show data — they explain it.

    Embedding Predictions

    Add Einstein Discovery predictions directly in your dashboards:

    1. Build a predictive model in Einstein Discovery
    2. Select the outcome you want to predict
    3. Einstein analyzes your data and creates the model
    4. Embed the prediction as a field in your visualization
    5. Users see predicted outcomes alongside actual data

    Use Cases

    • Sales dashboard with deal close probability
    • HR dashboard with employee attrition risk
    • Marketing dashboard with campaign ROI prediction
    • Operations dashboard with demand forecasting

    Configuring Tableau Pulse

    Setting Up Metrics

    1. Define your key metrics (revenue, users, conversion rate)
    2. Set the measure, time dimension, and filters
    3. Configure audience — who sees which metrics
    4. Set digest frequency (daily, weekly)

    What Pulse Delivers

    • Metric summary: AI-written plain language summary of current state
    • Trends: Automated trend detection (improving, declining, stable)
    • Anomalies: Unusual changes flagged automatically
    • Comparisons: How current period compares to previous
    • Contributors: What's driving changes in the metric

    Ask Data Best Practices

    • Add synonyms to your data source (e.g., "revenue" = "sales" = "income")
    • Use clear, descriptive field names
    • Add data source descriptions
    • Create curated data sources optimized for Ask Data queries
    • Train users on the types of questions they can ask

    Key Takeaways

    • •Embed Einstein Discovery predictions directly in Tableau dashboards
    • •Tableau Pulse delivers AI-written summaries, trends, and anomaly detection
    • •Configure metrics with audience targeting for personalized insight delivery
    • •Optimize data sources with synonyms and descriptions for better Ask Data results
    0 of 4 lessons marked complete

    Progress is kept in this browser only.