Cert Prep: Salesforce AI Specialist

What is a scatter plot, and why is it used in data analysis?

A scatter plot is a graphical representation of the relationship between two quantitative variables, where each point represents an observation. Scatter plots are used to visually assess the relationship between variables, helping to identify patterns, trends, or potential correlations.

After turning on the Lead Intelligence View, what additional step must an admin take for users to see it?

Add the Intelligence View button to the Lead List View in Setup.

How does a user return to the original list view from the Contact Intelligence View?

Click List View in the button menu.

How do data analytics software tools help solve complex business scenarios?

Data analytics tools scale data collection and visualize trends across large datasets, enabling businesses to identify patterns, optimize decision-making, predict outcomes, and uncover insights that guide strategy. These tools streamline processing vast amounts of data, making it easier to address complex problems and improve operational efficiency.

What's the difference between data-driven decision-making and traditional decision-making?

Data-driven decision-making is based on data analysis, while traditional decision-making is based on intuition and personal experience.

Which type of view is now available for leads, contacts, and accounts?

Intelligence. Intelligence View provides key metrics and actions for leads, contacts, and accounts in one place, improving efficiency in managing records.

Copilot Builder

Tool for managing a copilot's actions and settings, testing conversations, and auditing activity.

What two types of information shown on the Copilot Analytics Dashboard?

User Data and Action Data.

What are three Einstein Studio Features?

• Copilot Builder. • Prompt Builder. • Model Builder.

Toxicity Scoring

• Einstein Trust Layer scores content on toxicity. • Toxicity scores are logged and stored in Data Cloud as part of the audit trail.

What is meant by the term "linear relationship" in the context of correlation and regression?

A linear relationship means that the relationship between two variables can be represented by a straight line on a scatter plot. In a linear relationship, changes in one variable are consistently proportional to changes in the other variable, either positively or negatively.

What does an r-value of -0.52 indicate?

A modest negative correlation. An r-value of -0.52 falls within the range of -0.40 to -0.69, which is categorized as a modest correlation. The negative sign indicates the relationship is inversely proportional.

What does a negative r-value indicate in Pearson's correlation?

A negative r-value indicates a negative linear relationship between the two variables. As one variable increases, the other variable tends to decrease. The strength of this inverse relationship is determined by how close the r-value is to -1.

What does a positive r-value indicate in Pearson's correlation?

A positive r-value indicates a positive linear relationship between the two variables. As one variable increases, the other variable also tends to increase. The strength of this relationship is determined by how close the r-value is to 1.

From where is the Einstein Generative AI Audit and Feedback Data Report package accessed? A. Data Cloud. B. Marketing Cloud. C. Sales Cloud.

A. Data Cloud. Einstein generative AI audit and feedback data is stored in Data Cloud.

A company is preparing to reach out to potential leads who have shown interest in the company's latest product. The company wants to send personalized emails based on each lead's interactions and interests. Which feature should the company use? A. Einstein Sales Emails. B. Einstein Service Replies. C. Einstein Automated Contacts.

A. Einstein Sales Emails. Sales Emails uses data from Salesforce to generate email content that is tailored to the recipient's interests and previous interactions.

A marketing team wants to predict customer engagement with their upcoming campaign emails. They have historical engagement data, including open rates and click-through rates for past campaigns. What step must be taken when configuring a model in Model Builder to predict future customer engagement with campaign emails? A. Ensure the model is trained on a diverse dataset that includes both engaged and unengaged customer interactions. B. Configure the model to send emails automatically based on predicted engagement scores. C. Focus the model training exclusively on customers with the highest engagement rates to improve model accuracy.

A. Ensure the model is trained on a diverse dataset that includes both engaged and unengaged customer interactions. For a predictive model to accurately forecast customer engagement, it must be trained on a dataset that reflects the full spectrum of customer interactions, including both high and low engagement.

A Salesforce admin recently deployed Einstein Copilot. One user reported that a conversation with the copilot triggered an unexpected action. The admin wants to analyze the interaction between the user and the copilot to troubleshoot the issue. Where in the Copilot Builder can the admin review conversation data? A. Event Logs. B. Plan Canvas. C. Conversation Preview Panel.

A. Event Logs. Event logs help you monitor and troubleshoot the copilot's conversation activity with Salesforce users.

A company's marketing team wants to create a Flex prompt template that generates personalized event invitations based on the contact's interests and previous interactions with the company. How should the team ensure that the prompt template generates personalized invitations for each contact? A. Include instructions in the prompt template to use the contact's name and interests when generating an invitation. B. Manually input each contact's details into the prompt template before generating an invitation. C. Integrate the template with dynamic CRM data fields for automatic customization.

A. Include instructions in the prompt template to use the contact's name and interests when generating an invitation. Flex prompt templates in Prompt Builder offer a versatile and efficient way to automate the creation of text for PDFs, social media posts, newsletters, emails, record fields, and more with Einstein generative AI and the LLM (large language model).

New AI model architecture and availability of extensive training data are two factors in the rapid improvement of generative AI. What's the third? A. Increased parallel computing power. B. AI optimizing AI code. C. Larger data storage capacity of servers. D. Faster satellite data connections.

A. Increased parallel computing power. The rapid improvement of generative AI is driven by new model architectures, extensive training data, and the availability of increased parallel computing power, which allows AI models to process vast amounts of data more efficiently.

Which broad category would an AI system fit into if it's used to determine the optimal price of an airline ticket? A. Numeric prediction. B. Classification. C. Robotic navigation. D. Language processing.

A. Numeric prediction. Numeric prediction involves forecasting a numerical value based on historical data. In Salesforce AI, tools like Einstein Prediction Builder can predict outcomes such as optimal pricing by analyzing data patterns. This technique is useful for determining things like sales forecasts, pricing strategies, and revenue predictions.

A sales team wants to use Prompt Builder to quickly draft personalized emails for contacts or leads. Which prompt template type should the team use? A. Sales Email prompt template. B. Field Generation prompt template. C. Record Summary prompt template.

A. Sales Email prompt template. Sales Email prompt templates are designed to help sales teams draft personalized emails for contacts or leads.

A Salesforce AI Specialist is tasked with creating a predictive model to forecast sales revenue for the next quarter using Model Builder. The specialist decides to use historical sales data as the training dataset. Which step should the Salesforce AI Specialist take first when using the Model Builder to ensure the predictive model accurately forecasts sales revenue? A. Select "Create a model from scratch" and choose the historical sales data as the source. B. Increase the number of rows in the dataset to exceed 20 million for enhanced accuracy. C. Directly input the expected sales revenue figures for the model to predict.

A. Select "Create a model from scratch" and choose the historical sales data as the source. The initial step in creating a predictive model using the Model Builder involves selecting "Create a model from scratch" and choosing the appropriate data source, in this case, historical sales data.

What is one way the Einstein Trust Layer ensures data privacy? A. The Einstein Trust Layer detects and masks sensitive information before sending it to the large language model (LLM). B. The Einstein Trust Layer assigns role-based access controls to regulate data access. C. The Einstein Trust Layer enhances firewall protections to prevent unauthorized access.

A. The Einstein Trust Layer detects and masks sensitive information before sending it to the large language model (LLM). The Einstein Trust Layer uses data masking to prevent sensitive data from being exposed to the LLM.

A healthcare company is implementing Salesforce Einstein to enhance its customer service operations but is highly concerned about data privacy and healthcare regulation compliance. The company requires that no patient data is used for model training or product improvements. What feature of the Einstein Trust Layer addresses the organization's data privacy concerns? A. Zero-Data Retention Policy. B. Dynamic Grounding. C. Prompt Defense.

A. Zero-Data Retention Policy. The Zero-Data Retention Policy ensures that no data is used for LLM model training or product improvements.

What r-value indicates a very strong correlation?

An r-value between 0.90 to 1 or -0.90 to -1. The closer the r-value is to 1 or -1, the stronger the correlation. Values within these ranges are considered very strong correlations.

What is Salesforce Einstein Copilot? A. A platform for building and deploying custom AI models without coding, empowering businesses to leverage advanced predictive capabilities tailored to their specific needs. B. An AI-powered virtual assistant integrated into Salesforce, offering contextual guidance and automation to optimize user productivity. C. A tool in Salesforce that enables the creation, testing, customization, and management of AI-driven prompt templates, simplifying user tasks by integrating generative AI moments powered by CRM data.

B. An AI-powered virtual assistant integrated into Salesforce, offering contextual guidance and automation to optimize user productivity. Einstein Copilot is a customizable virtual assistant in Salesforce that enables you to bring the power of conversational AI to your business.

A support team manager wants to implement a feature that will help agents quickly catch up on ongoing customer conversations. The manager needs a solution that helps agents create an outline of completed conversations within a case, including the issue and resolution. Which feature meets these requirements? A. Einstein Service Replies for Chat. B. Einstein Work Summaries. C. Einstein Article Recommendations.

B. Einstein Work Summaries. Work Summaries provides real-time summaries of ongoing conversations, including the issue and resolution.

A Salesforce admin wants to ensure a prompt template dynamically generates personalized summaries of open cases for each account, using the most current data available in Salesforce. The prompt template must access and incorporate real-time account and case data. Which grounding technique should the admin use? A. Ground with Apex Merge Fields. B. Ground with Flow Merge Fields. C. Ground with Record Merge Fields.

B. Ground with Flow Merge Fields. Flow Merge Fields allow you to incorporate complex logic and real-time data into the prompt template.

A company aims to enhance its sales strategy by identifying the most influential factors leading to successful deals. The team is considering using Salesforce Einstein to analyze their sales data. Which scenario best justifies the use of Model Builder for this business requirement? A. If the team prefers to manually analyze sales data using spreadsheets without AI assistance. B. If the team wants to train a model on historical sales data to help identify key factors influencing successful deals. C. If the team is solely focused on generating weekly sales reports.

B. If the team wants to train a model on historical sales data to help identify key factors influencing successful deals. A model trained on historical sales data can help predict which factors will be most influential in closing future sales deals.

A Salesforce AI Specialist is tasked with improving lead qualification processes using Einstein Model Builder. The goal is to prioritize leads more effectively based on their likelihood to convert. Which consideration is most important when selecting data for training the model in Einstein's Model Builder for this purpose? A. Selecting only the data from the most recent sales quarter to ensure the model's predictions are up to date. B. Including data points from both converted and not converted leads to train the model. C. Using data exclusively from leads that have converted in the past, to focus the model on positive outcomes.

B. Including data points from both converted and not converted leads to train the model. A comprehensive training dataset that includes a variety of data points from both converted and not converted leads is crucial for the model to learn and differentiate the characteristics of leads that are more likely to convert.

What is a key benefit of grounding prompt templates with CRM data in Prompt Builder? A. It scores content on toxicity to prevent unintended or harmful outputs by the LLM. B. It provides the large language model (LLM) with context to create personalized responses. C. It automatically updates Salesforce records based on AI-generated responses.

B. It provides the large language model (LLM) with context to create personalized responses. Grounding prompt templates allows the LLM to generate personalized responses based on the specific context of the CRM data.

Which feature of the Einstein Trust Layer helps limit hallucinations and decrease the likelihood of unintended outputs? A. Dynamic Grounding with Secure Data Retrieval. B. Prompt Defense. C. Toxicity Scoring.

B. Prompt Defense. Prompt Defense refers to system policies that help limit hallucinations and decrease the likelihood of harmful outputs.

Grounding

Basing prompts on specific, relevant CRM data to ensure the AI's responses are contextually appropriate.

Universal Containers wants to increase productivity for sales representatives. How can Einstein Copilot help them meet these requirements? A. By generating automated reports. B. By calling customers on behalf of the representatives. C. By augmenting prospect research by summarizing CRM records.

C. By augmenting prospect research by summarizing CRM records. Copilot can help you boost productivity by summarizing lead, opportunity, and other CRM records.

A sales team wants to use AI to prioritize outreach to potential customers, based on their likelihood to convert. Which feature of Einstein for Sales Cloud should the sales team use? A. Einstein Opportunity Scoring. B. Einstein Activity Capture. C. Einstein Lead Scoring.

C. Einstein Lead Scoring. Lead Scoring uses AI to score leads by how well they fit the company's successful conversion patterns, allowing sales teams to prioritize their leads based on these scores

A customer support team manages a high volume of customer inquiries daily. The team wants to leverage generative AI to decrease the time it takes to draft and send responses to customers. Which feature should the company use? A. Einstein Case Classification. B. Einstein Call Summaries. C. Einstein Service Replies for Email.

C. Einstein Service Replies for Email. Service Replies for Email helps users draft and send personalized email responses to customers based on recommended Knowledge articles.

For AI training to be considered deep learning, what does its neural network need more of? A. Nodes. B. Weights. C. Layers. D. Inputs.

C. Layers. Deep learning refers to neural networks with multiple layers (often referred to as "deep" networks) that allow the model to learn complex patterns from data.

If you ask a generative AI what its favorite color is, and it responds "blue," this is an example of what? A. Sentience. B. Opinion. C. Prediction. D. Randomness.

C. Prediction. The text that a generative AI generates is really just another form of prediction. Gen AI predicts the sequence of words that are likely to have meaning and relevance to the reader.

A Salesforce admin is creating a custom action for Einstein Copilot that retrieves customer order tracking information. The admin wants to ensure that the action performs reliably. What step can the admin take to ensure this custom action functions correctly in conversations? A. Add multi-paragraph, detailed action instructions that include all possible uses of the action. B. Use generic language for the action name to cover a broad range of scenarios. C. Test the action in Copilot Builder and iterate on the refinement process for the action instructions and descriptions.

C. Test the action in Copilot Builder and iterate on the refinement process for the action instructions and descriptions. Testing actions and refining action instructions and descriptions is essential to ensure the copilot action meets your requirements.

What limits programmers from handcrafting algorithms to perform tasks we associate with human intelligence? A. Not enough memory in modern computers. B. Laws that prevent the creation of AI. C. The sheer number of rules to account for, many of which are unknown. D. Too little coffee, too little time.

C. The sheer number of rules to account for, many of which are unknown. Human intelligence involves too many complex rules, many of which are unknown, making it impractical to handcraft algorithms. Machine learning, like in Salesforce Einstein, learns patterns from data to handle complex tasks efficiently.

In the Account Intelligence View, how does a user view insights on an account?

Click the icon beside a record name to open the side panel and view all available insights.

r-squared value is also known as the:

Coefficient of Determination. The R-squared value, or coefficient of determination, measures how well the independent variables explain the variance in the dependent variable in a regression model. It indicates the proportion of the variance in the dependent variable that is predictable from the independent variables.

How does correlation differ from linear regression?

Correlation measures the strength and direction of a linear relationship, while linear regression predicts the value of the dependent variable based on the independent variable. Correlation focuses on the relationship between two variables, while regression is used to create a predictive model for one variable based on another.

What is it called when AI interprets everyday language? A. Slang Translation. B. Text-to-task. C. Intention Prediction. D. Natural Language Processing.

D. Natural Language Processing. Natural Language Processing (NLP) is the field of AI that focuses on enabling machines to understand, interpret, and respond to human language.

Why is it important to differentiate between correlation and causation?

Differentiating between correlation and causation is crucial because correlation only shows an association, not a cause-effect relationship. Mistaking correlation for causation can lead to incorrect assumptions about the relationship between variables, potentially leading to faulty decisions or interpretations.

Which feature allows an admin to set filters based on user fields so that specific actions are visible only to certain users?

Dynamic Actions. Dynamic Actions in Salesforce allows admins to set filters based on user fields, making specific actions visible only to certain users based on their profile, role, or other criteria. This enhances customization and user experience.

What can distort our understanding of artificial intelligence? A. Solar flares. B. An unclear definition of artificial. C. Fictional representations of AI. D. A narrow view of what constitutes intelligence. E. C and D. Fictional representations of AI and a narrow view of what constitutes intelligence.

E. C and D. Fictional representations of AI and a narrow view of what constitutes intelligence. Fictional representations of AI and a narrow view of intelligence can distort our understanding of artificial intelligence by creating misconceptions about its capabilities and limitations.

True or False: Developers must create their own large language models in order to add natural language processing to their applications.

False. Developers do not need to create their own large language models; they can leverage pre-built models like Salesforce's Einstein, OpenAI's GPT, or Google's BERT, which provide NLP capabilities that can be integrated into applications.

True or False: The values of weights and biases in a trained neural network usually have an obvious connection to the inputs.

False. In a trained neural network, the relationship between weights, biases, and inputs is typically not obvious, as these values are learned through complex optimization processes and are often difficult to interpret directly.

True or False: Ensuring data privacy and confidentiality is only important at the data collection stage.

False. Data privacy and confidentiality are important at every stage of the data lifecycle, not just during data collection.

True or False: A correlation of 0 means there is no relationship between the variables.

False: A correlation of 0 means there is no linear relationship between the variables. However, there could still be a non-linear relationship present.

True or False: In linear regression, X and Y values are interchangeable.

False: In linear regression, X (independent variable) and Y (dependent variable) are not interchangeable because the regression equation predicts Y from X.

True or False: Linear regression uses X and Y as interchangeable values.

False: In linear regression, X and Y are not interchangeable. The analysis results will change if X and Y are swapped because X is the independent variable and Y is the dependent variable.

True or False: Pearson's correlation can show causation between two variables.

False: Pearson's correlation only measures the strength and direction of the linear relationship between two variables. It does not prove causation.

What does the r-squared value indicate?

How well the model fits your observations. The r-squared value is a statistical measure that indicates how close the data points are to the regression line, reflecting the model's goodness of fit.

What does it mean if the r-value is close to 0?

It indicates a very weak or no linear correlation between the variables. An r-value close to 0 suggests that there is little to no linear relationship between the two variables being analyzed.

Which NLP technique uses the part of speech to more accurately find the root of a word?

Lemmatization. Lemmatization considers the part of speech to find the accurate root form of a word, unlike stemming, which only removes suffixes without context.

What Is Linear Regression?

Linear Regression shows the direction and strength of the relationship between two numeric variables, but unlike correlation, regression uses the best-fitting straight line through the points on a scatter plot to predict Y values from X values. With correlation, the values of X and Y are interchangeable. With regression, the results of the analysis will change if X and Y are swapped.

Linear Regression Versus Correlation.

Linear Regression: • Shows a linear model and prediction, predicting Y from X. • Uses r-squared to measure the percentage of variation explained by the model. • Does not use X and Y as interchangeable values (because Y is predicted from X). Correlation: • Shows a linear relationship between two values. • Uses r to measure the strength and direction of the correlation. • Uses X and Y as interchangeable values.

What's the main difference between machine learning and traditional programming?

Machine learning uses algorithms to learn from data and make predictions, while traditional programming follows a predefined set of rules.

The strength of the correlation of an r value of -0.52 can best be described as:

Modest Negative Correlation. An r value of -0.52 falls in the range of -0.40 to -0.69, which is categorized as a modest correlation. Since the r value is negative, it indicates a modest negative correlation.

In what ways have Neural Networks impacted NLP?

NLP has become faster and NLP has become more contextually accurate. Improved Accuracy: Neural networks have enhanced accuracy in tasks like text classification and machine translation. Contextual Understanding: Models like BERT consider context from both directions in a sentence. Handling Ambiguity: Neural networks disambiguate words with multiple meanings based on context. Human-like Text Generation: Models like GPT produce human-like text for applications like chatbots.

What is natural language?

Natural language is the language used by humans for everyday communication, such as English, Spanish, or Mandarin. It is characterized by its variability, complexity, and ambiguity, which makes it distinct from structured or formal languages like programming languages. In AI, Natural Language Processing (NLP) refers to the ability of machines to understand, interpret, and generate human language.

Copilot Planner Service

Orchestrates how a copilot launches actions during a conversation to accomplish a task for the user. The Copilot Planner Service in Einstein Copilot coordinates the execution of actions, ensuring that the right tasks are triggered at the appropriate moments during a conversation, allowing users to achieve their objectives efficiently. This service optimizes the interaction flow and enhances user experience by automating actions based on conversational context.

Why is it important to check for outliers when performing a correlation or regression analysis?

Outliers can dramatically affect the results of correlation and regression analysis, leading to misleading conclusions. Outliers are data points that deviate significantly from other observations. They can distort the correlation coefficient and the regression line, making the relationship between variables appear stronger or weaker than it actually is.

What is the term for finding the underlying structure of text in NLP?

Parsing or Syntactic Parsing. Syntactic Parsing (or parsing) involves analyzing the grammatical structure of text to determine how words relate to each other, helping to uncover the underlying structure in NLP.

Which of these best describes correlation? A. Pearson's correlation is a technique that can show causation. B. Pearson's correlation is a technique that can show how outliers affect relationships between variables. C. Pearson's correlation is a technique that can show whether and how strongly pairs of quantitative variables are related. D. Pearson's correlation is a technique that can show nonlinear relationships.

Pearson's correlation is a technique that can show whether and how strongly pairs of quantitative variables are related. Pearson's correlation measures the strength and direction of the linear relationship between two quantitative variables, but it does not imply causation.

What does Pearson's correlation measure?

Pearson's correlation measures the strength and direction of the linear relationship between two quantitative variables. Pearson's correlation coefficient (r) ranges from -1 to 1, where values close to -1 or 1 indicate a strong linear relationship, and values close to 0 indicate a weak or no linear relationship.

Prompt Templates

Predefined prompts that can be reused across different scenarios to maintain consistency in responses.

Prompt Resolution

Replaces each merge field in the prompt template with CRM data to help refine AI prompts.

Call Summaries

Sales Cloud feature for creating and sharing editable summaries of voice and video calls

Call Explorer

Sales Cloud feature that enables users to quickly gather information about voice and video calls.

What is the purpose of using scatter plots in correlation and regression analysis?

Scatter plots help visualize the relationship between two quantitative variables, showing trends and potential correlations. By plotting the data points, one can visually assess whether a linear relationship exists and how strong that relationship might be.

Einstein Service Replies

Service Cloud feature that generates email and chat responses based on knowledge-base data.

Work Summaries

Service Cloud feature that predicts and fills a summary, issue, and resolution after customer conversations.

What is GDPR?

The General Data Protection Regulation (GDPR) is a data privacy law implemented by the European Union (EU) to protect the personal data of EU citizens. It gives individuals more control over their data, requires businesses to be transparent about how they collect, store, and use personal information, and imposes strict penalties for non-compliance.

What is an advantage of doing data analytics in the cloud?

The benefits of not requiring deep knowledge of networking and operating systems and the ability to stop paying for infrastructure resources when they aren't needed.)

In a scatter plot used for linear regression, what does the best-fitting line represent?

The best-fitting line represents the line that minimizes the distance between all the data points and the line itself. This line is known as the regression line and is used to predict the dependent variable (Y) based on the independent variable (X).

What is the significance of the r-squared value in linear regression?

The r-squared value indicates how well the model fits the observations. The r-squared value, or coefficient of determination, measures the proportion of variation in the dependent variable that is predictable from the independent variable. An r-squared value of 1 means a perfect fit.

What is the significance of the slope in a linear regression equation?

The slope indicates the rate at which the dependent variable (Y) changes for each unit increase in the independent variable (X). The slope determines the steepness of the regression line. A steeper slope indicates a stronger relationship between X and Y, with more substantial changes in Y for each change in X.

What does the y-intercept represent in a linear regression equation?

The y-intercept is the value of the dependent variable (Y) when the independent variable (X) is zero. The y-intercept represents the point at which the regression line crosses the Y-axis. It is the starting value of Y before any influence from X.

True or False: AI is a technology that enables machines to learn and perform tasks that would normally require human intelligence.

True. AI enables machines to simulate human intelligence by learning from data, making decisions, and performing tasks that typically require human cognitive functions, such as problem-solving and pattern recognition​.

True or False: Data cleaning is the process of removing or correcting errors and inconsistencies in the data to improve its quality and accuracy.

True. Data cleaning involves identifying and correcting errors, inconsistencies, or inaccuracies in datasets to ensure high-quality, reliable, and accurate data for analysis and decision-making. This is a crucial step in data preprocessing.

True or False: A database of business names, zip codes, and market value would be an example of structured data?

True. Structured data is organized and formatted in a way that is easily searchable and analyzable, such as databases with specific fields like business names, zip codes, and market value.

True or False: ChatGPT is a generative AI model capable of producing text that closely resembles human language.

True. ChatGPT is a generative AI model capable of producing text that closely resembles human language. ChatGPT, based on the GPT architecture, is a generative AI model designed to produce text that mimics natural human language. It is trained on large amounts of data to understand context and generate coherent, relevant responses.

True or False: Predictive analytics tools may be used to analyze security logs to anticipate and block bad web requests.

True. Predictive analytics is commonly used in security to anticipate and prevent issues, such as blocking bad web requests.

True or False: Serverless technology enables organizations to focus on the analytics instead of server configuration.

True. Serverless technology abstracts the server management aspect, allowing organizations to focus on building and running applications without worrying about the underlying infrastructure.

In which scenario would you use linear regression instead of correlation?

You would use linear regression when you want to predict the value of one variable based on the value of another variable. Linear regression is useful for creating a predictive model, while correlation is mainly used to assess the strength and direction of the relationship between two variables without making predictions.

Zero-Data Retention Policy

Your data isn't retained by third-party LLMs. We partner with Open AI and Azure Open AI to enforce the zero-data retention policy. • No data is used for LLM model training or product improvements by third-party LLMs. • No data is retained by the third-party LLMs. • No human being at the third-party provider looks at data sent to their LLM.

Correlation refers to the:

direction (positive or negative) and the strength (very strong to very weak) of the relationship between two quantitative variables. Correlation measures how two variables move in relation to each other. A positive correlation indicates that as one variable increases, the other tends to increase as well, while a negative correlation shows that as one variable increases, the other decreases. The strength of this relationship is expressed by the correlation coefficient, which ranges from -1 (perfect negative correlation) to 1 (perfect positive correlation).

The Einstein Trust Layer is a collection of:

features, processes, and policies designed to safeguard data privacy, enhance AI accuracy, and promote responsible use of AI across the Salesforce ecosystem.

What are the four building blocks of Einstein Copilot?

• Copilot. • Actions. • The reasoning engine. • The large language model (LLM). Copilot: Acts as the conversational interface that interacts with users to assist with tasks and provide guidance. Actions: Allow users to execute tasks directly within the conversation, making the workflow more seamless. The Reasoning Engine: Powers decision-making and ensures intelligent responses by analyzing context and data. Large Language Model (LLM): Generates human-like text and understands context, leveraging natural language processing (NLP) capabilities.

Types of Prompt Templates

• Field Generation. • Flex. • Sales Emails. • Record Summary.

Audit

• Prompts, responses, and trust signals are logged and stored in Data Cloud. • Feedback can be used for improving prompt templates. • Provides pre-built reports and dashboards for analysis.

Dynamic Grounding with Secure Data Retrieval

• Relevant information from a Salesforce record is merged with the prompt to provide context. • Secure data retrieval of Salesforce data for grounding the prompt based on the permissions of the user executing the prompt. • Secure data retrieval preserves in place all standard Salesforce role-based controls for user permissions and field-level security when merging grounding data from your CRM instance or Data Cloud.

Data Masking

• Sensitive data is detected and masked before sending the prompt to the LLM. • Data masking supports multiple regions and languages. • You can select what must and must not be masked.

What are the three main classifications of data?

• Structured. • Unstructured. • Semi-Structured. Structured data is highly organized and easily searchable, like databases with specific fields (e.g., spreadsheets). Unstructured data lacks a predefined structure, such as emails, videos, and social media content. Semi-Structured data has elements of both, with some organizational properties but no strict format, like JSON or XML files.

Prompt Defense

• System policies help limit hallucinations and decrease the likelihood of unintended or harmful outputs by the LLM. • System policies can vary for different generative AI features and use cases.

How does data flows through the Einstein Trust Layer?

• The data in the form of a prompt, flows from CRM apps, through the Einstein Trust Layer, to the large language model (LLM), which we'll call prompt journey. • The LLM generates a response using the prompt, which we'll call response generation. • The generated response then flows back through the Einstein Trust Layer and back to the CRM apps, which we'll call the response journey.