CCNA AI Fundamentals Questions

Predicting a continuous numerical value, such as the optimal discount amount for a deal, is a regression problem. In the context of Einstein, this would use a regression model to learn from historical deal data and output a specific discount percentage or dollar amount, rather than a category or cluster.

Concept drift occurs when the relationship between features and labels changes over time. Old data may no longer represent current sales patterns.

Entity extraction (named entity recognition) identifies specific pieces of information such as dates, order numbers, and names from text.

Text classification assigns a category (e.g., billing, technical) to a text, enabling automated routing.

Salesforce's AI Trust Principles emphasize organizational accountability for AI-driven outcomes, ensuring that the organization takes responsibility for the decisions and impacts of its AI systems. This principle aligns with ethical AI deployment by requiring governance, oversight, and mechanisms to address unintended consequences, rather than shifting blame to the technology itself.

Defining intents is a core requirement for an Einstein Bot because intents represent the specific goals or tasks users want to accomplish, such as checking an order status or resetting a password. The bot uses Natural Language Processing (NLP) to map user utterances to these intents, enabling it to route conversations appropriately. Without clearly defined intents, the bot cannot accurately understand or respond to user requests.

The target variable for a time-based prediction model like Einstein Prediction Builder must be computed at a precise, consistent point in time—in this case, exactly 30 days from lead creation. If the target is computed at varying intervals, the model will learn incorrect patterns, as it cannot distinguish between leads that converted at 31 days versus those that never converted, leading to biased or invalid predictions.

High precision means most leads flagged as hot are indeed hot, but low recall means many actual hot leads are missed, causing lost opportunities.

Narrow AI is designed for specific tasks, while general AI (AGI) would have human-like broad capabilities, which does not yet exist.

Oversampling minority class, undersampling majority class, and using appropriate metrics (like precision/recall) are common approaches. Using accuracy is not recommended. Training on original data without adjustment will produce a biased model.

Overfitting means the model memorized training data and fails to generalize. Data leakage inflates training performance but not test performance.

The model's high accuracy on test data but poor performance in production is a classic symptom of data drift. In Einstein Prediction Builder, the model was trained on historical data that may not reflect current customer behavior patterns, leading to a mismatch between training and production distributions. This is not a model training issue but a data environment shift.

Quality labeled historical data ensures the model learns from accurate patterns. Relevant features (e.g., usage frequency, support tickets) improve predictions. A representative dataset avoids bias and ensures generalization.

The Einstein Score is a predictive lead scoring model that outputs a conversion probability (0 to 100), and the confidence score (0 to 1) indicates the model's certainty in that prediction. A confidence of 0.65 is below the typical threshold (e.g., 0.75 or higher), meaning the prediction is less reliable and should be manually verified before acting on it.

Precision measures the proportion of recommended items that are relevant (purchased). For product recommendations, precision is key.

Text classification assigns predefined labels to text, making it the right choice for categorizing support cases.

Einstein Activity Capture relies on Salesforce user licenses to associate email activities with records. If the sender or recipient is not a licensed Salesforce user, the system cannot link the email to a user identity, causing it to be skipped during capture. This is the most common reason for missing emails because the feature is designed to log activities only for licensed users.

Normalization scales features like 'AnnualRevenue' and 'NumberOfEmployees' to a comparable range (e.g., 0–1 or with zero mean and unit variance). Without normalization, a feature with larger numeric values (e.g., revenue in millions) would dominate distance-based calculations in models like k-nearest neighbors or gradient descent, causing the model to undervalue the smaller-scale feature. By normalizing, both features contribute equally to the model's learning process, which is essential for many machine learning algorithms used in Einstein.

Unsupervised learning can cluster leads based on similarities, providing initial scores without labeled outcomes.

Transparency in AI-generated outcomes is a core ethical principle, especially in Salesforce's Einstein platform, where users must understand how predictions (e.g., lead scoring or opportunity insights) are made. Salesforce provides tools like 'Why This Prediction?' to explain model outputs, ensuring trust and accountability. Without transparency, users cannot validate or challenge AI decisions, leading to potential bias or misuse.

Natural language processing (NLP) deals with understanding human language, including intent detection. Computer vision is for images, reinforcement learning for decision-making, and generative AI for content creation.

Einstein Reply Recommendations require the chat channel language to be set to a supported language (e.g., English, Spanish, French, German, Portuguese, or Japanese). If the language setting is unsupported or mismatched, the model cannot generate suggestions, even if permissions and data flow are correct. This is a common misconfiguration because the feature silently fails when the language is not recognized.

Narrow AI excels at specific tasks, while general AI would possess human-like cognitive abilities across many domains. General AI does not yet exist. The other options confuse definitions.

Predictive AI can forecast churn likelihood, and sentiment analysis can identify dissatisfied customers early.

Handling missing values and outliers is the most critical data preparation step for Einstein Prediction Service because the underlying gradient boosting models (like XGBoost) are sensitive to data quality issues. Missing values can introduce bias or cause the model to misinterpret patterns, while outliers can disproportionately influence split decisions, reducing predictive accuracy for churn scenarios.

Einstein AI in Salesforce Data Cloud provides machine learning capabilities that automatically analyze customer behavior patterns, such as purchase history and engagement metrics, to generate predictive audience segments. This enables marketers to target specific groups without manual rule creation, leveraging Data Cloud's unified data model and Einstein's propensity scoring.

Supervised learning uses labeled training data where the correct output is provided. Lead scoring with historical outcomes is a classic supervised learning task.

Transparency is a core principle of trustworthy AI because it requires that AI systems operate in a way that is open and understandable, allowing stakeholders to see how decisions are made. Salesforce's AI ethics guidelines emphasize transparency to ensure that users can trust the system's outputs and that any biases or limitations are visible.

Supervised learning can predict what a customer might buy next based on labeled data (past purchases). Unsupervised learning can cluster customers but doesn't directly generate recommendations. Reinforcement learning is for dynamic environments, and generative AI creates content, not recommendations.

Einstein Discovery includes a natural language generation (NLG) engine that automatically produces plain-English explanations of model insights, such as key drivers and predictions. This allows non-technical users to understand the output without needing to interpret raw statistical data.

Splitting data into train/test sets allows evaluation of generalization. Ensuring balanced representation of categories prevents bias towards majority classes.

When classes are imbalanced (few churners), accuracy can be misleading. Recall measures how many actual churners are identified, which is the key business concern.

Low confidence scores in Einstein Object Detection typically indicate that the model has insufficient or imbalanced training data. Einstein AI requires a minimum number of labeled records (e.g., at least 100 per class) and a balanced distribution across predicted values to learn effectively. Without adequate data size and diversity, the model cannot generalize well, resulting in confidence scores below 80%.

Augmenting training data with blurry images helps the model learn to handle various quality levels.

Bias in training data leads to unfair outcomes for certain groups. This is a fairness and bias issue.

The output shows an average response time of 350ms for the Einstein model, which is within the typical performance expectation of under 500ms for real-time AI inference in Salesforce. This indicates the model is responding efficiently without significant latency issues, making option C correct.

The error indicates that the 'OrderStatus' intent has fewer than the required minimum number of training phrases. Einstein Bot requires at least 5 training phrases per intent to ensure reliable natural language processing (NLP) model training. Adding at least 5 training phrases to the 'OrderStatus' intent satisfies this requirement and resolves the deployment error.

Overfitting occurs when a model learns training data too well, including noise, and fails to generalize to new data.

Bias in training data can lead to unfair or discriminatory outcomes, a key ethical concern in AI.

Prompt Builder supports conditional merge fields using IF statements, allowing dynamic inclusion of case subject, description, and last 3 comments only when the case priority is High. This approach ensures the prompt is concise and relevant, avoiding unnecessary tokens or manual edits.

When fallbackBehavior is set to 'USE_MODEL_KNOWLEDGE', the Einstein GPT grounding policy instructs the AI to fall back to its pre-trained (model) knowledge if the retrieved grounding data does not meet the configured relevance threshold. This ensures the AI still generates a response based on its internal training rather than returning no answer or ignoring the policy entirely.

Einstein Sentiment is a pre-trained model that may not accurately interpret sentiment in industry-specific contexts. Retraining the model with domain-specific training data (e.g., using Salesforce's Intent and Sentiment API with custom datasets) adjusts the model to recognize sentiment nuances in that industry, improving accuracy beyond the generic baseline.

Einstein Lead Scoring uses historical data and machine learning models to analyze patterns from past leads and their conversion outcomes. This allows it to assign a predictive score to each new lead, indicating the likelihood of conversion, which is more accurate and data-driven than manual scoring based on subjective criteria.

The Einstein Trust Layer policy configured to mask specific fields ensures that sensitive data is replaced with masked values before being sent to the AI model. This preserves data privacy while still allowing the model to generate predictions and insights. Feature importance insights remain available because they are computed from the masked data, not the original values.

Predictive AI uses historical data to forecast outcomes, making it ideal for opportunity scoring.

Precision focuses on the proportion of positive identifications that are correct; improving precision reduces false positives.

Supervised learning can use user purchase history as labels to predict what a specific user might buy, enabling personalization.

Hallucination is when a generative AI model produces content that is factually incorrect or nonsensical but presented as if it were true.

Predictive AI uses historical data to forecast outcomes. Regression (a type of predictive AI) predicts a continuous value like revenue.

Large gap between training accuracy and validation accuracy is a classic sign of overfitting. The model memorized training data and fails to generalize.

Testing an Einstein Bot with sample conversations before deployment allows you to validate the bot's intent recognition, dialog flow, and response accuracy in a controlled environment. This practice helps identify and fix issues with phrase matching, slot filling, and escalation paths, ensuring the bot performs reliably in production.

Excluding protected attributes like race or gender from the training data and ensuring the model does not use correlated proxies is essential to prevent bias in Einstein Prediction Builder. This approach directly removes the risk of the model learning discriminatory patterns based on these attributes, as the platform relies on the data provided and does not automatically enforce fairness constraints. Including such attributes or relying on built-in fairness would not guarantee unbiased predictions because the model could still infer protected characteristics from correlated features.

Einstein Send Time Optimization (STO) uses a machine learning model that analyzes each individual contact's historical email engagement patterns—specifically their past open and click behavior—to predict the optimal send time unique to that contact. This personalized approach ensures that each recipient receives the email when they are most likely to engage, rather than relying on aggregate or rule-based heuristics.

Accuracy is defined as the ratio of correctly predicted instances (both true positives and true negatives) to the total number of predictions. An accuracy of 0.85 means that 85% of the model's predictions (whether 'won' or 'lost') matched the actual outcomes in the dataset. This is a standard classification metric that evaluates overall correctness, not just one class.

The first step in improving Einstein Prediction Builder model performance is to review the training data for missing values and ensure relevant fields are included. Missing values or irrelevant fields can introduce noise and bias, directly degrading predictive accuracy. Einstein Prediction Builder relies on high-quality, complete training data to learn meaningful patterns, so data quality issues must be addressed before any other tuning steps.

The exhibit shows the bot repeatedly asking for the customer's account number without offering alternative identification methods (e.g., email, phone number, or name). This is the primary issue because Einstein Bot's conversational design should include fallback paths to handle cases where the user cannot provide the requested information, ensuring a smooth user experience and reducing drop-offs.

The prediction API returns a probability of 0.85 for the label 'High Value'. In machine learning classification, this output represents the model's confidence score—the estimated probability that the input instance belongs to the specified class. It is not a direct measure of conversion likelihood, model accuracy, or F1 score; it is the raw posterior probability assigned by the model to the 'High Value' label.

Bias in training data can lead to unfair denials; transparency ensures customers understand AI decisions. Accuracy is important but not primarily an ethical issue.

Supervised learning uses labeled data (historical outcomes) to predict future outcomes, making it ideal for lead scoring.

The model was likely trained on general data and does not generalize to domain-specific language. This is a domain adaptation issue, which can be due to training data not being representative.

Einstein Activity Capture requires that user emails be hosted on a supported platform (Gmail or Outlook/Exchange) because the feature uses server-side synchronization via APIs (Google Workspace APIs or Microsoft Graph) to automatically log emails and events into Salesforce. Without a supported email host, the service cannot connect to the mail server to capture activity data.

Predictive AI forecasts outcomes (churn, CLV) using historical data, while generative AI creates new content like text or images.

Text classification categorizes emails, intent detection identifies the purpose, and entity extraction can identify account numbers or products to aid routing. Sentiment analysis is about emotion, not routing. Generative AI creates responses, not needed for routing.

Recall (true positive rate) measures how many actual churners are correctly identified. High accuracy with low recall indicates the model is biased toward the majority class (non-churners). Precision focuses on how many flagged churners actually churn, F1 is a balance, and accuracy is already high.

Low recall means many high-value customers are missed, leading to lost opportunities to retain them.

High precision means few false positives (won deals predicted correctly), but low recall means many actual wins are missed. The sales team will likely miss out on some convertible opportunities.

The Einstein Relationship Intelligence permission set grants users access to the feature's insights and data processing capabilities. Without this permission set, users cannot view or interact with Einstein Relationship Intelligence data, even if the feature is enabled at the org level.

Narrow AI, also called weak AI, is specialized for one task.

Computer vision can process images of documents, and entity extraction (often part of NLP) identifies specific data points. Sentiment analysis is for emotions, intent detection for purpose, and generative AI for content creation.

Retraining the Einstein Opportunity Scoring model with the latest opportunity data, including discount information, allows the machine learning model to automatically learn the new patterns introduced by the sales team's discounting strategy. This aligns with the AI Associate principle that models must be retrained on current data to maintain accuracy when business processes change, without requiring manual intervention.

This is a text classification problem using supervised learning, where each ticket is assigned to a predefined category based on labeled training data.

Unsupervised learning can find patterns (e.g., customer segments) and supervised learning can predict likelihood to buy, both enabling recommendations.

Einstein Case Classification uses supervised machine learning to automatically categorize cases. The model learns from historical data, so it requires a sufficient number of closed cases (at least 500) with the correct category field populated to train effectively. Open cases are not used because the model needs confirmed outcomes to learn from.