Required Steps for Setting Up Einstein Prediction Builder
An admin is using Einstein Prediction Builder to create a model predicting whether a support case will be escalated. Which THREE steps are required during the prediction creation process?
Quick Answer
Selecting the object and the specific records to train on is a required step in Einstein Prediction Builder because the model has to be pointed at a defined, relevant dataset before it can learn anything, in this case the case records the admin wants to use to teach the model what an escalation looks like. That dataset scope works together with another required step, selecting the input features: once the training records are defined, the admin also has to choose which fields, like case description, account history, or product category, the model should actually consider as predictive signals. Neither step alone is sufficient; a well-chosen dataset with no meaningful features gives the model nothing to learn from, and well-chosen features applied to the wrong or overly broad set of records will teach the model the wrong patterns. Together, defining the object and records and defining the features are what let Einstein Prediction Builder build a model that's actually relevant to the outcome being predicted, here case escalation. This is a useful pattern to recognize generally: Prediction Builder setup questions tend to test whether you know that scoping the training data and selecting the predictive fields are both mandatory, distinct configuration steps rather than something the platform infers automatically.
⚠ Common exam trap
A common mix-up: candidates confuse the model creation steps with post-deployment integration tools like Einstein Copilot or Einstein Discovery, leading them to select options that are not part of the actual prediction creation wizard.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Select features (input fields) for the model
Selecting features (input fields) is a fundamental step in building a prediction model with Einstein Prediction Builder. These features are the independent variables that the model uses to learn patterns and make predictions about the target field (e.g., case escalation). Without selecting relevant features, the model cannot be trained effectively.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Run Einstein Discovery to validate the model
Why it's wrong here
Discovery is separate; Prediction Builder has its own validation.
- ✓
Select features (input fields) for the model
Why this is correct
Required: choose relevant fields like case origin, priority, etc.
- ✓
Select the prediction field (binary classification)
Why this is correct
Required: choose which field to predict, e.g., 'Escalated' (True/False).
- ✗
Configure Einstein Copilot to trigger the prediction
Why it's wrong here
Copilot is a separate feature; not part of Prediction Builder creation.
- ✓
Select the object and records to train on
Why this is correct
Required: define the dataset (e.g., all cases from last 6 months).
Go deeper
Related to this question
About these practice questions
This AI Associate question is part of Courseiva's 753-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →
Same concept, more angles
3 more ways this is tested on AI Associate
These questions test the same concept from different angles. Work through them to make sure you can recognise it however the exam phrases it.
Variation 1. An admin is building a custom AI prediction with Einstein Prediction Builder for a binary classification problem. Which THREE steps are required in the configuration? (Choose 3)
hard- A.Define a custom Apex class for data transformation
- ✓ B.Select a prediction field (the field to predict)
- ✓ C.Select the data set (records used for training)
- ✓ D.Select features (input fields for the model)
- E.Configure a trigger to retrain the model daily
Why B: The required steps are: select prediction field, select data set, and select features. Apex and triggers are not needed.
Variation 2. An admin is configuring Einstein Prediction Builder to predict case escalation. Which TWO components must be selected during setup?
hard- A.Prediction explanation template
- ✓ B.Prediction field (binary classification target)
- ✓ C.Features (input fields)
- D.Data set (records to train on)
- E.Prediction score field name
Why B: Einstein Prediction Builder requires a binary classification target field to define the outcome being predicted—in this case, whether a case will escalate. This field must have exactly two distinct values (e.g., 'Yes'/'No' or 0/1) to train the model. Without specifying the prediction field, the builder cannot determine what event to forecast.
Variation 3. An admin is setting up Einstein Prediction Builder to predict whether a lead will convert. The admin has selected the prediction field and data set. What is the next step in the configuration wizard?
hard- A.Train the model immediately
- B.Define the prediction explanation
- C.Choose the prediction score field
- ✓ D.Select features (input fields) to train the model
Why D: The Einstein Prediction Builder wizard proceeds: select prediction field, select data set, select features, define prediction field, then train.
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This AI Associate practice question is part of Courseiva's free Salesforce certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AI Associate exam.