Einstein Discovery and its stories are a powerful part of Tableau CRM allowing users to create point-and-click predictions. On this page, you will find a collection of blogs to help you understand what Einstein Discovery can do and how you can create predictions in your organization. If you want more make sure to check out this collection on assets.
- Define your use case
- Gather & prepare data
- Discover, analyze & strategize on your insights
- Create predictive models
- Deploy models & create AI-enhanced workflows
- Einstein Discovery in Tableau
- Manage your implementations
Define your use case
“Every great ML project starts with a great use case. ” Establish and clearly define the business goal. What business process do you want to improve? Who are the users and what will they do? Define what KPI will measure the goal. Envision what the AI-augmented workflows will look like.
Gather and Prepare data
Find data sources. Establish a means of labeling, cleaning, joining, and aggregating the data through Salesforce or other tools.
Discover, Analyze & Strategize on your Insights
Discover correlations and drivers in your data. Analyze it for insights. Create segmentations. Understand the conditions that create the greatest value. Strategize how these findings can be implemented into new experiences for your org.
Create Predictive Models
Create an instantiation of a model – to generate predictions for future events. Tune it until you are satisfied with its performance. Ensure you’re meeting your evaluation criteria. Dark launch the model in your org. Measure the difference between the predicted and actual values. Find and remove data leakage.
Deploy Models & Create AI-enhanced Workflows
Partner with your users to develop the interfaces and workflows that enhance their productivity – for both IC’s and management. Prototype the designs and implementation.
Einstein Discovery in Tableau
Deploy models specifically for integration in Tableau. Native integration to the world’s leading analytics platform, consume model outcomes in different ways.
Manage your Implementations
Following up on your models after they have been deployed.
Want to be notified when new blogs are published? Sign up below.