You are here:
Implementation Tasks for Predictive AI Solutions
The process of implementing predictive AI solutions typically involves a series of tasks.
Required Editions
| Available in: All Editions supported by Data 360. See Data 360 edition availability. |
Common Tasks
| # | Task | Description | More Information |
|---|---|---|---|
| 1 | Define the Outcome to Predict | Define an outcome that you want to improve. It’s most likely a key performance indicator (KPI) for your business. Most organizations have many outcomes that are candidates for improvement. Start with one outcome that’s significant in your business. Over time, you can develop and apply predictive AI models to many different outcomes concurrently. | |
| 2 | Assemble the Training Data | Design and populate the training data (in a Data Model Object, or DMO) to contain the historical data you want the model to learn from. The DMO contains tabular data that you normalize for model training. You can use Salesforce data as well as data that’s external to Salesforce. |
Prepare Your Training Data
|
| 3 | Create the Model | Based on the training data, create and configure a predictive model that tells Einstein how to analyze your data. Models provide the settings and preferences that Model Builder uses to generate insights. |
Create Predictive AI Models From Scratch
|
| 4 | Evaluate Model Quality | Assess performance and refine the predictive model. | |
| 5 | Predict and Improve Outcomes |
Predict outcomes and get prescriptions using the model you deployed. |
Get Predictions, Prescriptions, and Top Predictors
|
Iterative Refinement and Continuous Improvement
Implementation is an iterative process rather than a linear one. Model Builder is designed for rapid exploration, experimentation, and implementation. You learn as you go. Every step of the way, you use built-in feedback to check your results, review your assumptions, ask new questions, make adjustments, and try again. Add a column to your training data. Change model settings. Resolve alerts, refine thresholds, and filter data. As you fine-tune your approach, each improvement can help lead you to better operational outcomes.
