AI Associate Practice Test 46 — 15 Questions

Question 1

A company wants to use AI to automatically recommend the best next action for service agents during a live chat with a customer. Which Einstein feature should be enabled?

Accepted Answer

Einstein Next Best Action. Einstein Next Best Action, is correct because it is specifically designed to recommend the optimal next step for service agents during live interactions. It uses AI to analyze the customer context, conversation history, and business rules to surface the most relevant action, such as offering a discount or escalating the case, directly within the agent's console.

Answer

Einstein Article Recommendations

Answer

Einstein Bots

Answer

Einstein Reply Recommendations

Question 2

A healthcare organization uses Salesforce to develop an AI model for patient readmission prediction. They must comply with HIPAA regulations. The dataset includes patient names, addresses, medical record numbers, and detailed clinical notes. The data scientist plans to train a supervised model using historical readmission outcomes. What is the most important data governance step before model training?

Accepted Answer

De-identify all protected health information (PHI) by removing or masking identifiers.. HIPAA mandates that protected health information (PHI) must be de-identified before it can be used for model training without patient authorization. Removing or masking identifiers such as names, addresses, and medical record numbers ensures the dataset no longer contains individually identifiable information, allowing the organization to comply with the HIPAA Privacy Rule while still using clinical notes for predictive modeling.

Answer

Use only aggregated data that does not include any patient-level details.

Answer

Obtain written patient consent for every record used in training.

Answer

Store the data in a separate, encrypted environment with access controls.

Question 3

A company is preparing data for Einstein Prediction Builder to forecast lead conversion. They have historical data with fields like Lead Source, Industry, Number of Employees, and Converted (boolean). Which data preparation step is most critical?

Accepted Answer

Ensure data completeness by handling missing values in Lead Source. Handling missing values in Lead Source is critical because Einstein Prediction Builder requires complete, high-quality data to train accurate predictive models. Missing categorical fields like Lead Source can introduce bias or cause the model to ignore important patterns in lead conversion. Ensuring data completeness through imputation or removal of incomplete records is a standard data preparation step for AI/ML in Salesforce.

Answer

Mix data from all lead sources without normalization

Answer

Use only the last 3 months of data for training

Answer

Remove all records with outliers in Number of Employees

Question 4

A company wants to use Einstein Bots to handle customer inquiries. They need to train the bot to understand different customer intents. Which TWO components are essential for defining bot understanding?

Accepted Answer

Intents. Intents (A) are essential because they represent the customer's goal or purpose behind a message (e.g., "Check Order Status" or "Reset Password"), and Einstein Bots use training phrases tied to each intent to classify incoming utterances via natural language understanding. Entities (B) are also essential because they extract specific variable data from within an utterance—such as order numbers, dates, or product names—that the bot needs to fulfill the matched intent. Together, intents and entities form the core NLU model that defines how the bot understands what the customer wants and the key details involved. Actions (C) are not part of defining understanding; they define what the bot does after understanding. Topics (D) and Dialogue Flows (E) govern conversation routing and the sequence of dialog steps, not the intent/entity understanding model itself.

Answer

Actions

Answer

Topics

Answer

Dialogue flows

Question 5

Which THREE are requirements for enabling Einstein features in a Salesforce org?

Accepted Answer

An eligible Salesforce edition. Option A is correct because Einstein features (such as Einstein Prediction Builder, Einstein Discovery, and Einstein Bots) are only available in eligible Salesforce editions like Enterprise, Performance, Unlimited, and Developer, so the org must be on a qualifying edition. Option C is correct because users need the appropriate permissions—such as the 'View Einstein Predictions' or Einstein feature-specific permission sets—to access and view predictions, since access is not granted by default. Option E is correct because Einstein's machine learning models require a sufficient volume of relevant, quality data (for example, a minimum number of records with historical outcomes) to train and produce meaningful predictions. Option B is not required because Einstein features work with standard and custom objects alike; creating custom objects is not a prerequisite. Option D is not required because Einstein features are enabled through Setup's Einstein settings and permission sets, not by activating an 'Einstein API' toggle.

Answer

Custom objects must be created

Answer

Activation of Einstein API in Setup

Question 6

A Salesforce admin is creating an Agentforce agent in Agent Builder to handle order status inquiries. The agent needs to look up order data from a custom object 'Order__c' and respond with the status. Which THREE components must the admin configure in Agent Builder?

Accepted Answer

Prompt instructions that tell the agent how to respond. In Agent Builder, the admin must configure a topic (D) such as 'Order Status' because topics define the scope of a conversation and group the instructions and actions the agent can use for a given subject. Within that topic, the admin must add prompt instructions (A) that tell the agent how to respond, since these instructions guide the agent's tone, behavior, and response format when handling order status inquiries. The admin must also configure an action (C) that queries the Order__c object, because Agent Builder uses actions (often backed by Flows, Apex, or prompt templates) to retrieve the actual order data the agent needs to answer the question. Options B and E are not required components: intents and entities belong to classic NLU models rather than Agent Builder's topic-and-action design, and routing to a human agent is a separate escalation concern, not a mandatory component for looking up and responding with order status.

Answer

Intents and entities for natural language understanding

Answer

A Flow to route the conversation to a human agent

Question 7

A service manager wants to automatically categorize incoming cases based on their description. Which Einstein feature should be used?

Accepted Answer

Einstein Case Classification. Einstein Case Classification uses natural language processing to automatically predict and populate standard case fields such as Type, Priority, and Reason based on the case description. It is purpose-built for categorizing incoming cases using historical data patterns. This directly matches the manager's goal of auto-categorizing cases from their descriptions.

Answer

Einstein Reply Recommendations

Answer

Einstein Vision

Answer

Einstein Article Recommendations

Question 8

A company is preparing customer data for a predictive model. They notice that many records have missing values for the 'annual income' field. Which approach is best to handle this issue while minimizing bias?

Accepted Answer

Use model-based imputation considering other features.. Model-based imputation (Option B) is best because it uses relationships between features (e.g., education, job role) to predict missing 'annual income' values, preserving data distribution and minimizing bias. This approach avoids the distortion caused by simple mean/zero imputation and retains sample size better than deletion.

Answer

Remove all records with missing values.

Answer

Replace missing values with the mean.

Answer

Set missing values to zero.

Question 9

A sales operations manager wants to use Einstein Opportunity Scoring to improve win rates. They want to view the opportunity score and understand why a particular score is high or low. Where can they see the score and explanation in Salesforce Lightning? (Choose TWO)

Accepted Answer

In the Opportunity list view as a column. Option C is correct because Einstein Opportunity Scoring adds a sortable 'Opportunity Score' field that can be displayed as a column in an Opportunity list view, letting the manager scan and rank opportunities by score directly from the list. Option D is correct because the Opportunity record page includes the Einstein Scoring component, which shows the score along with the top positive and negative factors (the explanation of why the score is high or low). Option A is incorrect because the Mobile 'Today' view surfaces prioritized items but is not where the score and its factor-level explanation are presented. Option B is incorrect because the Activity Timeline shows past and upcoming activities, not Einstein score factors. Option E is incorrect because Einstein Discovery dashboards are for building and viewing predictive models/stories, not the standard place to see an opportunity's score explanation.

Answer

In Salesforce Mobile App under 'Today'

Answer

In the Activity Timeline

Answer

In the Einstein Discovery dashboard

Question 10

A sales operations team wants to improve forecast accuracy by using AI. They currently use manual rollups. Which TWO Einstein features can help achieve this?

Accepted Answer

Einstein Forecasting. Einstein Forecasting (A) is correct because it uses AI to generate predictive forecasts from historical opportunity data, replacing manual rollups and directly improving forecast accuracy for the sales operations team. Einstein Opportunity Scoring (B) is correct because it applies AI to score opportunities based on their likelihood to close, giving reps and managers data-driven insight that sharpens forecast inputs and prioritization. Einstein Activity Capture (C) is not correct here because it only syncs emails and calendar events into Salesforce; it captures activity data but does not itself produce forecasts or opportunity scores. Einstein Discovery (D) is not correct in this context because it is a broader analytics and predictive-modeling tool for building custom models on datasets, not a purpose-built forecasting or opportunity-scoring feature. Einstein Lead Scoring (E) is not correct because it scores leads rather than opportunities, so it does not address forecast accuracy for the sales pipeline.

Answer

Einstein Activity Capture

Answer

Einstein Discovery

Answer

Einstein Lead Scoring

Question 11

For a real-time AI application that requires low-latency access to customer interaction data, which storage solution is most appropriate?

Accepted Answer

In-memory data store.. In-memory data stores (e.g., Redis, Memcached) store data in RAM rather than on disk, providing sub-millisecond read/write latencies essential for real-time AI applications that need immediate access to customer interaction data. This eliminates disk I/O bottlenecks and enables high-throughput, low-latency data retrieval for time-sensitive inference or decision-making.

Answer

Flat files on a network drive.

Answer

Relational database with complex joins.

Answer

Data lake with batch processing.

Question 12

Which TWO of the following are limitations of Einstein GPT? (Choose two.)

Accepted Answer

It requires structured prompts for best results. Option A is correct because Einstein GPT, like other large language models, performs best when given structured, well-formed prompts; vague or unstructured input often yields generic or off-target output, making prompt engineering a real limitation. Option C is correct because generative AI models can hallucinate and can reproduce biases present in their training data, so Einstein GPT may produce biased or factually inaccurate content that requires human review. Option B is not a limitation but a capability, since Einstein GPT can generate account summaries from CRM data such as leads. Option D is false because language support is uneven, with some languages performing far better than others. Option E is false because these models are trained on large datasets and typically require grounding with CRM data or fine-tuning for domain accuracy.

Answer

It can automatically generate account summaries from leads

Answer

It supports all languages equally

Answer

It requires no training data

Question 13

Which TWO statements correctly describe predictive AI compared to generative AI?

Accepted Answer

Predictive AI is used for forecasting numeric outcomes like churn probability. Option B is correct because predictive AI models (e.g., regression, classification, gradient boosting) are trained on historical data to output numeric or categorical forecasts such as churn probability, lead scores, or sales estimates. Option C is correct because generative AI (e.g., LLMs, diffusion models, GANs) is designed to create novel content — text, images, audio, or code — rather than predict a labeled outcome. Option A is wrong because lead scoring is a classic predictive classification task, not a generative one. Option D is wrong because the two paradigms use fundamentally different objectives and architectures (discriminative/predictive models vs. generative models like transformers or diffusion networks). Option E is wrong because producing realistic customer service responses is a generative AI capability, not a predictive one.

Answer

Generative AI is best suited for lead scoring

Answer

Both predictive and generative AI rely on the same algorithms

Answer

Predictive AI can generate realistic customer service responses

Question 14

An e-commerce company uses AI to provide product recommendations. The model suggests popular items but fails to personalize for individual users. Which type of learning could improve personalization?

Accepted Answer

Supervised learning. Supervised learning is correct because personalization is fundamentally a prediction problem: given a user's past behavior (features), predict what they will engage with or buy (label). The model is trained on labeled historical data — clicks, purchases, ratings — mapping user/item features to known outcomes, which is exactly the supervised paradigm. Popularity-based recommendations ignore user-specific signals; supervised models learn per-user preferences from those labeled interactions, enabling individualized recommendations.

Answer

Generative AI

Answer

Unsupervised learning

Answer

Reinforcement learning

Question 15

Which TWO actions are essential for ensuring transparency in an AI system? (Choose two.)

Accepted Answer

Log all AI decisions and allow audit. Option B is correct because transparency requires that an AI system's decisions be traceable and reviewable; logging all AI decisions and enabling audit creates an immutable record that lets stakeholders verify how and why outputs were produced. Option D is correct because providing clear explanations for AI decisions makes the system's reasoning intelligible to users, auditors, and regulators, which is the core of explainability and a key pillar of transparent AI. Option A is incorrect because hiding the model's internal logic directly undermines transparency, even if it protects intellectual property. Option C is incorrect because training on the largest dataset improves performance or coverage, not transparency. Option E is incorrect because obtaining consent from data subjects addresses privacy and lawful processing, not transparency of AI decision-making.

Answer

Hide the model's internal logic to protect intellectual property

Answer

Train the model on the largest dataset available

Answer

Obtain consent from all data subjects