Scenario practice questions

An AI system used for medical diagnosis has been shown to have lower accuracy for certain ethnic groups. The development team is considering releasing it anyway because most patients are from the majority group. Which ethical principle is most compromised?

Trap 1: Transparency

Transparency is about explaining decisions, not performance equity.

Trap 2: Accountability

Accountability is about who is responsible, not the ethical breach itself.

Trap 3: Privacy

Privacy is not directly related to group performance disparity.

A
Transparency
Why wrong: Transparency is about explaining decisions, not performance equity.
B
Fairness
Unequal performance across groups violates fairness.
C
Accountability
Why wrong: Accountability is about who is responsible, not the ethical breach itself.
D
Privacy
Why wrong: Privacy is not directly related to group performance disparity.

Question 2hardmultiple choice

A company uses Einstein Prediction Builder to predict which leads will convert. They have a binary outcome field 'Converted__c' which is true for 8% of leads. After training, the model shows high accuracy (95%) but very low precision for the positive class. What is the most likely cause?

Trap 1: The prediction field is not a binary field

The field is binary (true/false) as stated.

Trap 2: The prediction score field is not configured correctly

Prediction score field configuration does not affect model training accuracy/precision.

Trap 3: The dataset is too small for training

Small dataset may cause overfitting, but the described symptoms point to class imbalance.

A
The prediction field is not a binary field
Why wrong: The field is binary (true/false) as stated.
B
The data is imbalanced favoring the negative class
Imbalanced data causes the model to predict majority class most of the time, yielding high accuracy but low positive precision.
C
The prediction score field is not configured correctly
Why wrong: Prediction score field configuration does not affect model training accuracy/precision.
D
The dataset is too small for training
Why wrong: Small dataset may cause overfitting, but the described symptoms point to class imbalance.

Question 3mediummultiple choice

A service manager wants to analyze historical case data to identify the most common reasons for escalations and get actionable suggestions to reduce them. Which Einstein tool should they use?

Trap 1: Einstein Prediction Builder

Prediction Builder predicts binary outcomes, it does not provide diagnostic analysis or suggestions.

Trap 2: Einstein Case Classification

Case Classification automatically assigns field values; it does not analyze reasons or suggest improvements.

Trap 3: Einstein Conversation Insights

Conversation Insights analyzes call recordings, not historical case data.

A
Einstein Prediction Builder
Why wrong: Prediction Builder predicts binary outcomes, it does not provide diagnostic analysis or suggestions.
B
Einstein Case Classification
Why wrong: Case Classification automatically assigns field values; it does not analyze reasons or suggest improvements.
C
Einstein Discovery
Discovery analyzes data, generates stories, and offers improvement suggestions and operational prescriptions.
D
Einstein Conversation Insights
Why wrong: Conversation Insights analyzes call recordings, not historical case data.

Question 4mediummulti select

A service team wants to use Einstein GPT to improve agent productivity. Which TWO Einstein GPT features are designed specifically for service scenarios?

Trap 1: Einstein Lead Scoring

Lead Scoring is a sales feature.

Trap 2: Sales GPT - Email Generation

This is a sales feature, not service.

Trap 3: Einstein Opportunity Scoring

Opportunity Scoring is a sales feature.

A
Einstein Lead Scoring
Why wrong: Lead Scoring is a sales feature.
B
Service GPT - Case Summaries
Case summaries are a core Service GPT feature.
C
Service GPT - Knowledge Article Drafts
Article drafting is a Service GPT feature.
D
Sales GPT - Email Generation
Why wrong: This is a sales feature, not service.
E
Einstein Opportunity Scoring
Why wrong: Opportunity Scoring is a sales feature.

Question 5hardmultiple choice

A company uses Einstein Bots to handle customer service inquiries. The bot often fails to understand complex requests, leading to escalations. Which improvement strategy is most effective?

Trap 1: Route all complex requests directly to human agents without bot…

This undermines the bot's purpose.

Trap 2: Increase the confidence threshold for intent matching to avoid…

Higher threshold may cause the bot to not understand more queries.

Trap 3: Reduce the number of dialogue options to simplify the bot's logic.

This limits the bot's capability.

A
Train the bot with additional intents and example phrases for complex scenarios.
More training data improves NLU accuracy.
B
Route all complex requests directly to human agents without bot interaction.
Why wrong: This undermines the bot's purpose.
C
Increase the confidence threshold for intent matching to avoid misclassification.
Why wrong: Higher threshold may cause the bot to not understand more queries.
D
Reduce the number of dialogue options to simplify the bot's logic.
Why wrong: This limits the bot's capability.

Question 6mediummultiple choice

A retail company uses Einstein Prediction Service to forecast customer churn. To improve model accuracy, which data preparation step is most critical?

Trap 1: Select only the top three features based on correlation.

Feature selection is secondary to data quality.

Trap 2: Use a different algorithm like neural networks.

Algorithm choice is less impactful than data quality.

Trap 3: Increase the dataset size by collecting more customer records.

More data without quality can degrade performance.

A
Select only the top three features based on correlation.
Why wrong: Feature selection is secondary to data quality.
B
Clean the dataset by handling missing values and outliers.
Proper data cleaning ensures the model learns accurate patterns.
C
Use a different algorithm like neural networks.
Why wrong: Algorithm choice is less impactful than data quality.
D
Increase the dataset size by collecting more customer records.
Why wrong: More data without quality can degrade performance.

Question 7mediummultiple choice

An administrator is configuring Einstein Lead Scoring. After activation, lead scores are visible in the lead record page. However, some leads that are clearly not interested (e.g., bounced email) are scored 90+. What is the MOST likely reason?

Trap 1: Einstein Lead Scoring only works for imported leads

It works for all leads.

Trap 2: The lead score field is not added to the page layout

The field is visible, so layout is fine.

Trap 3: The model requires at least 2000 converted leads to be accurate

Even if accurate, the issue is about training data quality, not quantity.

A
Einstein Lead Scoring only works for imported leads
Why wrong: It works for all leads.
B
The lead score field is not added to the page layout
Why wrong: The field is visible, so layout is fine.
C
The administrator did not exclude bounced leads from the training population
Training data should include leads that had a chance to convert; excluding bounced leads prevents the model from learning from irrelevant records.
D
The model requires at least 2000 converted leads to be accurate
Why wrong: Even if accurate, the issue is about training data quality, not quantity.