Salesforce AI Associate Exam Set 1

Preventing Bias in AI Systems

• To prevent bias when training an AI system, use diverse and representative datasets.

Handling Data Inconsistencies

• When inconsistencies are identified in a dataset for an AI model, investigate and apply data quality techniques.
• This ensures training data is clean, accurate, and consistent.

Impact of Poor Data Quality

• Poor data quality leads to unreliable and inaccurate model predictions and generations in predictive and generative AI models.

Data Quality Dimensions

• Inconsistent data entry methods, like using a text field in one region and a picklist in another for the same data, affects data consistency.
• Inconsistent data entry makes uniform analysis difficult.

Evaluating Data Quality

• When evaluating the quality of sales data, first identify business objectives to determine important data quality dimensions.

Avoiding Bias in AI Models

• To avoid introducing unintended bias in an AI model, omit demographic data, as it can introduce bias based on sensitive or protected attributes.

Enforcing Consistency Across Accounts

• To enforce consistency, implement naming conventions or predefined user-selectable values, which will ensure consistent data entry.

Salesforce’s Trusted AI Principle of Empowerment

• Salesforce's Trusted AI Principle of Empowerment aims to empower users of all skill levels to create AI solutions easily, using low-code tools.

Salesforce’s Trusted AI Principle of Transparency

• Salesforce’s Trusted AI Principle of Transparency entails a clear and understandable explanation of how and why AI makes decisions.

Impact of Duplicate Records

• Duplicate records negatively impact data quality, leading to confusion and inaccuracies in reports and processes.

AI in CRM for Sales Representatives

• AI within CRM provides call summaries to give sales representatives insights into past conversations and helps them follow up effectively.

Predictive AI vs. Generative AI

• Predictive AI predicts outcomes based on data, while generative AI creates new and original content based on learned patterns.

AI Tool with Web of Connections

• Neural networks, which mimic the human brain, are the AI tool that is a web of connections, guided by weights and biases.

How Cloud Kicks Enhances Sales with AI

• Cloud Kicks should use Einstein Lead Scoring and Case Classification to improves sales targeting and streamlines support by predicting case fields.

Salesforce’s Trusted AI Principle of Inclusivity

• Salesforce’s Trusted AI Principle of Inclusivity aligns with testing AI with diverse and representative datasets to serve all users fairly.

Einstein Functionality for Self-Service

• Einstein Bots within its functionalities guide customers through solutions in real time in a guided self-serve application

Assessing Data Quality for Einstein Prediction Builder

• Leverage data quality apps from AppExchange when preparing for Einstein Prediction Builder for a fast and effective assessments.

Trusted AI Principle in Consequence Scanning Workshops

• Conducting Consequence Scanning Workshops aligns with the Trusted AI Principle of Inclusivity by considering the impacts of AI on different groups.

Impact of Data Quality on AI Trustworthiness

• High-quality data improves the reliability and credibility of AI-driven decisions.

User Consent in AI Data Privacy

• Without explicit user consent, AI can violate data privacy principles.

Impact of Assessment on Business Outcomes

• Assessing data quality sets a standard for prediction performance and reliability.

Benefits of AI Personalization

• AI personalization leads to greater satisfaction and customer loyalty.

AI Applications for Sales and Support

• Lead scoring, opportunity forecasting and data classification directly supports sales and service optimization.

Data Excluded by Salesforce

• Salesforce excludes demographic data to mitigate bias in Marketing Cloud Einstein.

How Incomplete Data Impacts Product Recommendations

• Incomplete data leads to less accurate AI recommendations.

Improving Prediction Quality

• To improve prediction quality using more data, focus on accuracy.

Salesforce Feature for Removing Duplicates

• Duplicate Management helps detect and merge or prevent duplicate records.
• It is a Salesforce feature to remove duplicate customer records.

Correct Field to Capture Preferred Name

• Use a Text field to allows flexible entry of any name format when recording a customer's preferred name.

Immutable Traits Causing Bias

• Immutable traits like financial status can introduce bias when used improperly in AI.

Benefit of Human-AI Interaction

• Combining human judgment with AI insights leads to better decisions, which is a key benefit of effective human-AI interaction.

Supporting AI Transparency

• Providing rationale and model cards ensures decisions are explainable and transparent.
• This supports AI transparency to provide on-platform access to prediction rationale and model cards.

Safety Guideline for Generative AI

• Creating guardrails to mitigate toxicity and protect PII supports safety guidelines for generative AI.

How AI Assists in Lead Qualification

• AI evaluates customer data to score and prioritize leads.

Key Data Quality Consideration

• High-quality data is critical for effectively training and fine-tuning AI models.

Action Introducing Bias

• Using a dataset that underrepresents perspectives introduces bias in training data, leading to biased predictions and unfair outcomes.

Impact of AI Data Quality

• AI systems amplify bias, which is a consequence of low-quality data.
• Low-quality data leads to AI learning and repeating biased or incorrect behaviors.

AI Feature for Sales Reps

• Call Summaries automatically capture key conversation points to streamline logging to quickly and accurately log calls.

Principle of Least Privilege

• Limiting access reduces risk of misuse or exposure of sensitive data, and protects data.

Example of Ethical Debt

• Releasing flawed AI solutions builds up ethical debt, requiring future corrections.

Societal Implication of Excluding Ethics

• Ignoring ethics can lead to biased AI that negatively affects vulnerable populations.

Al evaluates customer data

• AI evaluates customer data to score and prioritize leads