Get Certified in less than 18 minutes

Key Points:

• The Salesforce AI Associate and AI Specialist certifications are free to take until the end of 2025
• The AI Associate Certification focuses on general knowledge about AI, and as such is suitable for anyone to take, even people who aren't very familiar with Salesforce.
• AI Fundamentals - Study the Glossary of terms
• AI Capabilities in the CRM - Know about Salesforce's Einstein AI Products
• Ethical Considerations of AI - Know Salesforce's Trusted AI Principles
• Data Considerations for AI - Review the Data Quality Metrics here and this article on data bias

Resources:

• Exam overview (Salesforce)
• Exam guide (Salesforce)
• Trail Mix (Salesforce)
• Generative AI Glossary (Salesforce)
• Salesforce Trusted AI Principles (Salesforce)
• Data Bias (Salesforce)
• Data Considerations for AI (Salesforce)
• 20 practice questions (Saas Guru)
• 40 Practice questions (Dineshyadav)

Transcript

00:00:00 hi everyone my name is Alex diamond and in this video I'm going to take you through everything you need to know to pass the Salesforce AI associate exam let's get into it this exam was the first AI focused exam released by Salesforce AI in particular agent force at the top of everyone's Minds at the moment and will be for some time to come sales force has recently made both this and the AI specialist certifications free for anyone to take up until 2025 so now is the perfect time to get certified

00:00:31 okay so we're going to go through six sections in this exam guide we're going to go through the exam and what's on it then we're going to talk about the fundamentals of ai ai capabilities in the CRM ethical considerations of AI data for AI and further resources that you can use to study for it go through the exam guide now I'll be pulling this up from November 2024 um if you're watching this video in the future then it makes sense to make sure that what's being covered is still on the exam unlike most sales for certifications

00:00:59 this is a of general knowledge you don't need to have much sales force knowledge to pass this certification I'd normally advise for anyone who is doing any certifications to make sure that they spend the time to go through every single sales force trail head badge um to get a grasp of the product but because this is more general knowledge it's likely that you've came across some of the terms and Concepts that we're going to go through in this video in other places um a lot of these modules will

00:01:26 come up again uh when you do the AI specialist certification anyway so so look it's not going to hurt if you go through and do the badges but don't feel that you need to do them to pass this exam I would suggest that you do the ones that come up on the AI specialist certification when you're going for that exam for this exam itself uh you need a mark of 65% to pass so you need 26 marks out of 40 there's only three answers per question so you'll have a pretty good chance of getting it right um especially

00:01:53 if you can eliminate one wrong answer which brings you to 50% of the time so you know if there isn't a question that comes up on the exam that you've got no idea what the right answer is just spend some time to try to eliminate what you think the wrong answers are there are four sections on the exam the first is AI fundamentals in this you need to explain basic principles and applications of salili the second is AI capabilities in the CRM now this is worth only 8% so it's the smaller section of the exam in

00:02:28 it you need to be able to identify CRM AI capabilities and describe the benefits of AI as they apply to the CRM that CRM is Salesforce ethical considerations of AI you need to describe the ethical challenges of AI for example human bias in machine learning lack of transparency Etc and apply salesforce's trusted AI principles to given scenarios so one of the things I would advise anyone to learn um before they sit this exam is sales forces trusted AI principles and we'll go through through those in depth

00:03:01 in this video series ethical considerations of AI is the highest L section of this exam it's worth 39% so definitely learn those trusted AI principles finally we've got data for AI which is describing the importance of data quality and describing the elements and components of data quality okay section one of the exam is AI fundamentals and in this section you're going to be asked to explain the basic principles and applications of AI within Salesforce and differentiate between the types of AI and their

00:03:35 capabilities but in the early days of AI but it's moving incredibly quickly if anyone sort of looks back at AI in 2020 and where it's came there's no doubt that it's moving at a rapid Pace when you think of AI you're probably thinking of what you see in the movies so the Terminator movie you have the Terminators themselves and Skynet the Terminator let's say it's at a similar level of intelligence to humans um this is an example artificial general intelligence artificial general intelligence is when AI is at a similar

00:04:05 level of intelligence to humans you've also got Skynet in The Terminator movies and that was smarter than humans much smarter um it invented a way for to send the Terminators back in time so Skynet is an example of an artificial super intelligence where a long way from artificial general intelligence and artificial super intelligence some predictions are sort of putting it about 40 years away before we reach artificial general intelligence and not long before we reach an artificial super intelligence what we're talking about

00:04:34 when we talk about AI today is artificial narrow intelligence this is where we use it for a narrow range of uses as the name would suggest and give it a particular set of prompts to generate an outcome that where focused on so there's several different types of AI uh you've got predictive AI which is where it's Mak making predictions Based on data something like Einstein lead scoring or opportunity scoring analytic AI when it's used to analyze something so Einstein forecasting or call insights and generative AI when it's used to

00:05:06 generate text or images such as chat GPT agent force or co-pilot there's a glossery of terms that Salesforce has put together surrounding Ai and I encourage you anyone to read through it we're going to go through some key terms to remember here as well so you've got natural language processing or NLP which is a field of artificial intelligence that combines computer science and Linguistics to give computers the ability to understand interpret and generate human language in a way that's meaningful and useful to

00:05:35 humans computers don't understand all the nuances of data that humans do so they need to pass this data to understand it so passing when we talk about passing is breaking the input data down into smaller chunks that the NLP can classify and thus understand an algorithm so that's turning an input into an output for a series of rules such as you know maths 1+ 1 equals 2 algorithms used by AI can be quite complicated structured data structured data is data that's structured in a clear way so I think spreadsheets are a

00:06:09 classic example or CRM objects unstructured data so this is data that isn't structured in a particular way and isn't ready for classification unstructured data might be a website or an image machine learning it's a way to have computers learn from data with minimal programming instead of writing code you feed a machine data and it builds its own logical function based on this data neural network a machine learning model that makes decisions in a similar way to the human brain it uses Mass to analyze

00:06:38 many different interconnected data points generative AI you're using AI to generate something based on a prompt text image videos this going be a summary a prediction a translation anything and this is where a lot of people go to when they think of AI today you've also got computer vision or robot Vision so that's where you're analyz an image using AI this could be used to scan and classify documents um and we're also seeing it being used for things like Medical Imaging to analyze brain scan by terms AI capabilities and CRM so

00:07:12 in this section you're going to identify the CRM AI capabilities and describe the benefits of AI as they apply to CRM this section is worth the least amount of marks when you're thinking about this section also think Einstein um because the exam was written before agent force became agent force and also before Co pilot was talked about a lot there's many different types of AI that exist within Salesforce so youve got predictive AI which is used to make predictions Based on data such as Einstein lead scoring or opportunity

00:07:41 scoring analytic AI which is used to summarize something so forecasting or call insights generative AI used to generate text and images such as chat jpt agent force co-pilot um AI assistance which can be Voice or natural language based Einstein functionality which is embedded in various Salesforce apps like sales cloud service Cloud can help businesses to discover new insights predict the outcomes of things based on past data recommend next best actions automate routine tasks and also generate tailored

00:08:15 content like an email or code this can be helpful to many different business areas so for sales it can help you to better understand customers leading to an increase in conversions and sales agents also know who to spend time focusing on and can use generative AI to get a summary of the account generate a close plan generate a follow-up email call coachings available which um we'll go into in a bit more detail in the AI specialist exam service you can use AI to decrease response time through AI assisted service channels

00:08:43 case classification field generation and response generation you can use AI for marketing to write emails and improve email open rates with Einstein send time optimization you can use it to generate segments analyze campaign Effectiveness you can also use it to work out who might be be getting fatigued by the amount of emails you're sending them and actually pullback on the number of messages that you're sending to them as well for Commerce Cloud you can use it to increase Revenue by showing people

00:09:10 the best products for them um or the products that they're most likely to buy you can also use it to generate product descriptions and personalized search let's talk about some of Sal Force's AI tools you got Einstein bots so don't confuse this with agent force Einstein Bots allow to build a smar system in channels including chat m messaging and voice they focus on routine inquiries like password resets and providing status updates to improve response times while also reducing the need for additional staff if a chatbot

00:09:40 can't resolve an inquiry then the customer can also be directed to a real agent Einstein prediction Builder this is a simple pointand click wizard that allows you to make predictions on your non-encrypted Salesforce data you can get this in the form of a binary classification EG yes or no or a number figure um such as lead scoring or it's on a scale of 0 to 100 with 100 being great and zero being terrible next best action helps your sales and service teams predict the next best action to take based on what worked

00:10:09 for other accounts and cases so it's looking at your case and account data and then working out based on that what other people did and what worked and what didn't work Einstein Discovery helps you to analyze data and Einstein generative AI such as agent force or copilot is used to generate a field summary of an object and emails to similar fashion to chat JP ethical challenges of AI let's talk about the ethical challenges of AI which is a big section on this exam um so you'll be asked to describe the ethical

00:10:45 challenges of AI for example human bias in machine learning a lack of transparency Etc and also to apply salesforce's trusted AI principles to given scenarios so AI can only spit out answers based on the data that it's being trained on data sources are often biased for instance people might have only been looking for certain things when they created the data set and captured the data set the sample might be representative of only a particular group or trait and not the wider population at large the data might be

00:11:13 sourced from the internet which is full of people's biases when you're looking at data it's important to prioritize intent over demographic attributes so for instance you might be targeting demographic targeting ads to a specific interest group as opposed to an age group or a geographic Lo region and that helps you to eliminate bias responsible use of AI means trying to eliminate any potential bias in the data before releasing it into the wild um it's also about using AI responsibly so does what

00:11:42 you are creating have the potential to cause harm and if so how can you mitigate that risk um you shouldn't falsely represent AI as a person people should give consent for their data to be used and know what it's going to be used for they should also get a clear benefit from the use of their data UL ultimately you don't want AI to come off as intrusive or manipulative you want it to be something that gives people value this next slide uh we're talking about salesforce's trusted AI principles um I

00:12:11 want you to remember these because you will get a whole heap of questions about these on the exam and this is one of the few things that you do need to memorize to do well at this exam sales force is trusted AI principles the acronym you can use to remember them is ratei r a t eii responsible are accountable a transparent T empowering e inclusive I rtii responsible accountable transparent empowering and inclusive one of salesforce's trusted AI principles is that it is responsible so safeguarding

00:12:47 human rights and protecting the data that we're entrusted with accountable seeking and leveraging feedback for continuous Improvement transparent delivering a transparent user experience to guide users through machine driven recommendations being transparent about how Salesforce is building its AI empowering promoting economic growth and employment for our customers their employees and society as a whole Empower our customers to use our AI responsibly and inclusive respect the societal values of all those impacted not just

00:13:18 the creators I'd encourage you to read this blog article as well meet salesforce's trusted AI principles it'll help to give some more context behind the uh trusted AI principles as well too okay so data for AI is another big section it makes up 36% of the AI associate exam there are two components to this one is describing the importance of data quality the other is describing the elements or components of data quality let's talk about some of the data considerations for AI um so again this is something that is worth

00:13:56 memorizing so talking about this here there's an article have linked to as well um but from it youve got the age of data the completeness of data accuracy of data consistency of data duplication of data and the usage of data set so let's go through each of those now age of data when was the data last updated when was it gathered age is a key consideration here old data is likely to be skewed um or may not be reflective of the current situation completeness data our key data needed for a model present on all records so a good thing

00:14:35 to think about here is how often are the fields filled in if you're using AI to analyze industry for instance you would need that industry field on the account to be filled out in most places for it to be considered complete accuracy is your data accurate is it from a trusted Source um do you trust the source of your data consistency is the format spelling and language consistent across all records duplication you may have duplicate records in your data which is going to skew the data as well and usage

00:15:08 of data sets is your data being used for anything or is it being just entered into the CRM so um you want to make sure that the data that you're analyzing is intended to be used you're not just capturing a whole heap of data points that are going to be useless and aren't going to be used by you look into Data you can use the data quality analytics dashboard app to audit data quality in Salesforce now this is something that available on the app exchange talk about data biases now as well and there's another blog article

00:15:37 that I'll I'll link to here um data bias to determine if data is biased you can look at where the data was sourced where it comes from is it first or third party is it being filtered is it being categorized and labeled um is a human involved in categorizing and labeling that data and if so would they be biased when is the data captured so how fresh or recent is your data older data can lead to inaccurate answers you can also think about who the data uh represents so is it representative of all of your customers

00:16:12 and users or are you just using a subset of data is there more data on one group than another are you getting information about different people in the data set what uh so certain types of data can introduce biases such as race ethnicity country of origin gender age sexual orientation religious of political affiliation you got to ask yourself does it make sense to include these in your model um and also also are there proxy values there uh like post code which can skew based on income data leakage um so this is an

00:16:45 interesting one and what we're talking about here is the data point coming after use so for instance rain and umbrella sales and umbrella sale isn't a good indicator that it's going to rain but it is more likely to occur after rain whereas rain would be a good indicator of umbrella s so you just got to sort of understand that correlation may not mean causation in all cases further resources I'm now going to bring up some links uh that'll be in the comments of the video as well that you can also use to help study and prepare

00:17:20 for the AI specialist exam thank you and I hope you enjoyed this video