Loading...
technologybusinesscareers
Making Data Cloud Understandable for Admins
About this Episode
Today on the Salesforce Admins Podcast, we talk to Abhishek Saxena, Technical Architect at Copado.
Join us as we chat about how he learned Data Cloud and why understanding context is the key to making Agentforce shine.
You should subscribe for the full episode, but here are a few takeaways from our conversation with Abhishek Saxena.
Overcoming the complexity of Data Cloud
As a developer and architect, Abhishek isn't lacking for technical knowledge about the Salesforce platform. But even he found it hard to get his head around what Data Cloud was and what it could do.
Abhishek attended community events, scoured LinkedIn posts, studied videos, and even read a book about Data Cloud. But there were so many new terms being thrown around, and he still couldn't explain the difference between a data lake object, a data model object, and a data source object.
"Even though there was a lot of buzz around Data Cloud and how it is such an amazing, innovative solution," Abhishek says, "I was not able to grasp what it does in an easy fashion." Luckily, he had an ah-ha moment that helped him see the big picture, and so he's giving a presentation at Dreamforce to share what he's learned.
What Data Cloud actually does
Abhishek's talk, "A Beginner's Guide to Data Cloud," will get you up to speed in 20 minutes or less. As he explains, the main thing to understand is that Data Cloud is about data unification.
If you have your data in a bunch of different places, you used to have to dedicate significant developer time to maintaining APIs that allowed Salesforce to share information with your other platforms. With Data Cloud, you have everything on one record, with Salesforce and Slack as the front door. You have a complete 360 view of your customer, regardless of where the information is.
Why Data Cloud is crucial for Agentforce
Getting a complete picture of your customers is doubly important when it comes to Agentforce. AI agents are extremely context-dependent: they do a much better job when you "ground" them with extra parameters.
As Abhishek says, "If you give agents good data, your responses are going to be much more personalized and better." Data Cloud allows you to give your AI agents a much more specific picture of your customers, opening the door for better and more effective automations.
If you're coming to Dreamforce, make sure to come to Abhishek's presentation so you can be a Data Cloud pro. And don't forget to subscribe to the Salesforce Admins Podcast so you never miss an episode.
Podcast swag
Learn more
-
Salesforce Admins Blog: From Pillars to Lakes: Using Data Cloud As Your Source of Truth
Admin Trailblazers Group
Social
Full show transcript
Mike:
So have you ever tried to figure out what a data lake is and then ended up wondering where the lifeguard's at? Ditto. Today on the Salesforce Admins podcast, we're talking with Abhishek Saxena, a Salesforce consultant with a developer's mindset and a teacher's heart. Abhishek's going to take us through his journey of learning Data Cloud and how it went from something buzzworthy to something he could really explain to a five-year-old. So if you've ever felt overwhelmed by new tech or just really weren't sure where to begin this episode's for you. Plus Abhishek gives us a sneak peek at his Dreamforce session and why understanding context is key in making AI tools like Agentforce shine. So with that, let's get Abhishek on the podcast. So Abhishek, welcome to the podcast.
Abhishek:
Thank you so much for having me, Mike.
Mike:
I'm excited to talk about this because of all the things going on at Salesforce, we've got a big event coming up in just a few weeks from when this is going to air, and you're doing a presentation there, but before we talk about that, let's find out a little bit about you. How did you get started doing stuff in the Salesforce ecosystem and want to present at Dreamforce?
Abhishek:
Certainly. So I have been working in the Salesforce ecosystem for about 10 years now, it'll be 10 years later in November. I started off immediately after my college where I was studying computer science engineering. I always had an affinity to computers and how they work, so it was an easy choice to what to study.
But getting into Salesforce, that was a happy coincidence. My hometown, where I'm originally from, it's called Jaipur, it's in India, and that's where I did my engineering from as well, Jaipur is traditionally not touted as a tech hub, but for some divine reasons there were several Salesforce consultancies that were trying to make it big in that area when I was just graduating, I got an offer to work for one of them as a Salesforce consultant after a series of intense grueling interviews. But yeah, that's how I got started, and I have never looked back since then.
Mike:
Fresh out of college. And let's see, people heard this in the past, I've recently had somebody point out to me that my Salesforce experience is old enough it could graduate from high school. So your Salesforce experience is somewhere in middle school, I guess, at this point.
Abhishek:
Yeah, sounds about right. I'm eager to get to high school and get to be the cool kid.
Mike:
Well, right now it's getting its first iPhone and being popular. I don't know what ten-year-old kids do nowadays, I don't have kids, so we'll just move on. Eventually my experience will be in college and that'll scare me.
Abhishek:
Yeah, I hope I can get the certified technical architect or MVP or one of those cool badges to show around everyone.
Mike:
Yeah, I don't know. I mean, being in the ecosystem is just fun enough too. It's not going down the road with all those biker patches or like a race suit. It's not, what was that game? Pokemon? You don't have to collect them all. But you're talking Data Cloud, so tell me from your experience, what's so interesting about Data Cloud that admins need to know about it?
Abhishek:
So as I mentioned, I've been in the ecosystem for a long time and got decent exposure on Sales Cloud, Service Cloud, Experience Cloud, so yeah, just the core platform in general. But over the last few years, Data Cloud has been one of those top things, and most talked about things from Salesforce at events everywhere. So even though there was a lot of buzz around Data Cloud about how it is such an amazing innovative solution, it could replace some of the other tools that an organization might be using, I was not able to grasp what it does in an easy fashion. Especially around two years back, 2023 Dreamforce when it was relaunched after being renamed from Genie, I was trying to understand what it does, what could be its use cases, maybe look at some Trailhead modules, but at that point in time, there wasn't a lot that was available.
So Trailhead modules weren't baked out completely. There weren't any demo environments that I could get hands on, and it just felt a little bit overwhelming. I knew my friends who were doing Marketing Cloud, Einstein, CDP, that sort of stuff, knew what was going on, but us commoners who were just working on the core clouds.
Mike:
Us commoners.
Abhishek:
Yeah, we weren't privy to that information as much as they were. So I started to get down on this journey to learn a little bit myself because yes, seems like everyone was talking about it. So I started by attending different community events in the States and Canada, trying to attend the Data Cloud specific sessions, read people's LinkedIn posts, watch some videos. I even read a book on Data Cloud-
Mike:
Wow.
Abhishek:
... that one of our friends in the community had put out, but even then it was just so overwhelming trying to think about it logically. I am an engineer, I like to think about things in a logical analytical manner that, okay, I'll first learn the alphabets, then I'll start to learn words, sentences, so on and so forth. So I wasn't able to get that topography down in my mind that way to get started with Data Cloud. There was these few workshops or actually Salesforce Data Cloud trainings that Salesforce teams were doing several times a month. I had to attend that one training probably four times until it finally clicked to me that, okay, now it feels that I have the basics down of what's going on with Data Cloud, and if someone comes up to me and asks me about it, I'll probably be able to give at least a satisfactory answer on what exactly is Data Cloud.
But at this point, now I had several months of experience trying to learn Data Cloud, I thought that, okay, let me go ahead and do the Data Cloud certification as well just to quiz myself if I know enough. I did that and I passed. So that was the validation I was seeking that, okay, now I am somewhat of a Data Cloud consultant. But I wanted to share this journey with everyone else because I consider myself fairly technical with a consultant and a developer background, and if it took me a lot of time to get through, then I would not blame anyone else in the ecosystem who's not as technical to not completely understand that, Hey, why Data Cloud? What is Data Cloud? Why is this important? So that's why I thought that yes, I need to share my knowledge with everyone else.
Mike:
No, I get that. And what's so relevant about all that is when companies roll out features, and I'm not just speaking for Salesforce, but anybody, there's often a lot of thought and process that has gone into it, and just because you don't get it right away within a sentence or two or 10 seconds into somebody's presentation, that's not on you. That's just speaking to the complexity of perhaps the situation, where you're at in your learning journey, and then also just the complexity of the product that you're trying to understand. I think it's a lot for you to say, it took me a few months to kind of wrap my head around this. It doesn't just happen right away. And I think some people get really frustrated in themselves when they're like, well, I just don't get it.
Abhishek:
No, absolutely. I agree with you. And yeah, I have been very lucky that I like what I do. So even after work, I'm not afraid to put in some hours to just sharpen my skills. But everyone's situation is not the same, you've got a life outside work as well, you've got so many things going on that sometimes you're just not able to put in those 10,000 hours to get perfect on something. And that's why I feel that if you attend my session on A Beginner's Guide to Data Cloud, at least in those 20 minutes that we are together, I could give you a good framework about, okay, after you leave it that I understand what is Data Cloud. I know what are some key use cases and why everyone says that it's needed for Agentforce. So that's what I'm trying to get out in those 20 minutes for everyone who has not yet had the opportunity and the luxury to actually focus on going through the Data Cloud trail certification modules or just having the training on the job of their own organization adopting Data Cloud yet.
Mike:
Yeah. What was it for you, as you were investigating looking into and learning about Data Cloud, what was the first kind of piece of it that started to click for you that really was like, oh, I think I'm starting to understand it?
Abhishek:
Yeah, great question. So when I was learning about it, and as I said, Data Cloud is a slightly different beast than the regular core clouds like Sales, Service, Experience, it's something called as a near core cloud. So it's not directly built on Salesforce, it's Salesforce adjusting, you can say. So some of the concepts that were being introduced, they were not getting my other knowledge that I had gained through learning Sales, Service and Experience Cloud. There were concepts called as Data Lakes, there were some activations, segmentations, things that I had never really learned in the other clouds while working on it. So even though it was somewhat difficult to get the vocabulary for Data Cloud to get started, once I was able to work through those first few basic things, the biggest aha moment for me was when I saw that Data Cloud can help you consolidate a lot of different records.
So I'll try to explain this and I go into much more detail in my session as well. If anyone in our listeners is in the attendance, I would love to have you in my session. But just to give you a recap or a quick preview of what I'm going to talk about is that let's say your company uses Salesforce, your company also uses SAP for something else. Your company maybe uses Trello for tracking some stuff as well. So you have your record your identity as, since my name is Abhishek, all these systems have my identity there in SAP or ServiceNow, Abhishek can do certain things, maybe that has my personal email for some reason, or maybe a different phone number. And then the Abhishek on the Salesforce ecosystem, that record, that contact or user, has slightly different details. Maybe the address I put in is different. Maybe the phone number I put in is my work phone, and in the other one I put in my personal phone.
So what Data Cloud ends up doing is that from these different identities of Abhishek that exist in these separate platforms that are not linked, it allows you to create one record where all of your identities are linked. So in case I want to get Abhishek's email in SAP, I can just call out to Data Cloud and Data Cloud will give that to me instead of myself having to write three or four different APIs trying to de-duplicate records, trying to find some common way or how it's called an external key of linking that, okay, this Abhishek in Salesforce and that Abhishek in SAP are the same person.
Data Cloud takes all of that complexity out of this and just tells you that, okay, Abhishek's record exists in five systems, here are the values of different things in those five systems. And I could just get that. So once I understood this, I was like, wow, this unified data that Data Cloud is giving you could have so many ramifications. And that's again, one of those things that I explored in that presentation that how agents that need this sort of information can benefit from Data Cloud.
Mike:
You were so close. I was about to say, wow, you just described Data Cloud and you didn't talk about agents. But no, what's crazy is thinking back in my umpteen hundred years of Salesforce experience, there's been different waves of where's your data? For a while it was, well, Salesforce is CRM, and then they had Service and then that takes care of that. But there's another system for this and there's another system for that, another system for that. And the world just kind of operated that way. It was like just understood, people had multiple systems to log into. And I remember for a long time when I was a Salesforce administrator, the big question was, well, such and such system, does X, can't Salesforce do that? And there was a big push for admins and developers to kind of rebuild all of this functionality into Salesforce.
And so you would have these really robust time management or project management apps and the platform could more than easily handle it. And you didn't have to be a coder, you could build a project management app with a few objects and a couple dozen fields and some relationships, and there you're ready to go. And then it kind of became like, well, wait a minute, why are we spending all this time building on this platform when we can just connect things? But there was no sense of unification there. And I think in your description, what I heard is finally now you can have an entire enterprise where the front door is Salesforce or Slack and it can access all of that data and you're getting a complete view of a customer regardless of where the system information is, and that allows other applications and other things to run and do their job just well, but you still get that unified pull back in, and then you don't have the burden of a developer maintaining 200 APIs on the back end.
Abhishek:
100%, Mike, you got a spot on.
Mike:
See, you're a good teacher. Look at what you just taught me.
Abhishek:
Well, you are not a bad student either.
Mike:
Well, I think a lot of people have had Data Cloud explained to them in terms of agents and Agentforce, and there's power there, but there's also the first step out the front door, which is, yeah, but what if your customer service people or your salespeople sit down and actually get that full view that we've been talking about regardless of where the system of that information lives?
Abhishek:
Certainly, and that reminds me actually, so the terminology that Data Cloud uses is that you get the customer 360 view. Meaning that, okay, your organization uses Salesforce for CRM, you are using Azure to actually log into different systems, single sign on, maybe Excel and other Azure products. You're using Workday for HR related things. You're using Asana for project tracking. And like I said, you have different identities in these different systems, different values for maybe same fields, similar fields like email address, regular address, phone number, et cetera. But with Data Cloud, you are able to unify all of these different identities of a single user and then present that within your Salesforce systems as the customer 360 where you can access all these different details and take much more intelligent decisions. So hopefully now our listeners are able to get that, okay, why is it so important for Agentforce and why is it so important in general.
Mike:
Yeah. So then that's my next question. Why is it so important for Agentforce?
Abhishek:
So agents, like any AI, and I'm sure our listeners must have used some sort of AI at this point, ChatGPT, Gemini or anything. When do agents do their best work? They do their best work when they have good context to go on and a good prompt that the user asks. Now, what is context? Context means that if you ask it to write me a sales email, if that's your prompt, it's going to give you a very generic answer, a generic response that, okay, Dear ABC, are you interested in our sales services? Are you interested in buying our product? But that's going to be it. It's going to imagine a few things, it's going to skip out some important details. So basically the answer is going to be generic and not as helpful for your particular organization's scenarios.
Now, if you provide the agent with good context that whenever agent responds, it knows that I'm working for Coral Cloud Resorts, I am responsible for handling bookings, I can refund things, I can access customer details and which objects to look at, where to get the information from. If your agent has these informations, or again, a simpler example that if you are in ChatGPT and tell it that, okay, I want to write a sales email as a customer service agent for Coral Clouds, my name is Abhishek Saxena, when you start your email, start with the salutation. This is intended to be for the guests who are staying at the resort. And if I give all of these details and then hit enter, the response I'm going to get is going to be a lot more personalized.
And that's how you can automate it to send to any user that is using your agent. That's what I mean by why context is so important for any AI to work. And with Data Cloud, as we have discussed in the podcast so far, it gives you that unified profile. It gives you the customer 360 so that your agents have all the information about your particular user who is asking the question or to wherever you're trying to use it. Because good data, if you give it good data, your responses are going to be much more personalized and better.
Mike:
And that's a lot of what we do in the Agentforce Now workshops like the one I do, the intro one where we talk about just prompts and even just writing a better, more descriptive prompt. If you think about it, I kind of reference it in the same way of it's like learning to cook with one cookbook or learning to cook with a library of cookbooks. If you have more references then you have more ways of looking at things and more context into providing better information and a better response.
Abhishek:
Certainly.
Mike:
I love it. I'm curious to know, in learning something, there's always something you're going to run into that maybe frustrates you or maybe for me, just bounces off the surface. What was it about Data Cloud that was kind of something you really had to work to try and understand that maybe other admins might be having difficulty understanding as well?
Abhishek:
Yeah, great question. Like I said, most of my experience was on Sales Service Experience Cloud, so I was not a data specialist or a data consultant. So wrapping my head around what is a data lake? What is a data warehouse? What is the difference between a data lake object versus a data model object versus a data source object? How are they getting stored? Where are they getting stored? Is it not just another Salesforce object in which we are storing everything?
So that's sort of basics of data processing that a typical data engineer knows, but a typical Salesforce consultant, admin developer does not, that was somewhat difficult for me to wrap my head around because in my training on working on Salesforce platform for so many years, I have been accustomed to think about data in terms of objects, Socket queries, and that form. But with Data Cloud, it's a little bit different. So that was one of the things that threw me off initially, and I had to look at some resources even outside of Salesforce's Data Cloud trainings to understand that data lake is not actually a lake. Data warehouse is not a big factory in which there are tons of books with data. So yes, that was something that threw me off initially.
Mike:
Yeah, those terms when they first came out, I know I've done some podcasts like with Skip Sauls, and he's talked about putting a cabin next to the data lake and stuff that. I'm like, okay, who comes up with these terms? It's not helping me understand. Where's the data pond? Can you have a data swimming pool? Somebody decided lake.
Abhishek, this has been a fun conversation. I think it's going to be a very compelling presentation at Dreamforce, but also I think it's been very inspirational, at least for me, that even the hard stuff is still not that hard to learn. I go back to a saying, everything in the world is created by somebody at some time, and if that somebody can figure it out, you certainly can too. So Data Cloud and Data Lakes was created by somebody, which means they figured it out, so it can't be that hard for anyone else to figure it out.
Abhishek:
Yeah, no, absolutely. And as I said, you got to learn the alphabet before you can learn to work with words and work with sentences. And I firmly believe that if you cannot explain something to a five-year-old, then you don't understand it yourself. And again, a five-year-old needs to have that structure to understand it. So hopefully with the session that I've put together, everyone can learn the alphabet of Data Cloud and even start to make their own sentences.
Mike:
I like it. Well, we'll see how many people can describe it to a five-year-old after we're done. Thanks so much for coming on the podcast. This was a lot of fun.
Abhishek:
The pleasure has been mine, Mike. Thank you for having me.
Mike:
Okay, big thanks to Abhishek for joining us and breaking down what it really means to learn Data Cloud. I think I picked up a few things in just that short bit of time. Now, whether you're prepping for a certification or just trying to make sense of new tech, I really think his story was a reminder that persistence pays off and that even you as a Salesforce admin don't need to have to figure it all out in one day.
So be sure to catch his session if you're going to dream for us and share this episode with a fellow Salesforce admin maybe who's also struggling to understand some new technology or just getting started with Data Cloud. Either way, until next time, we'll see you in the cloud.
Hosts & Guests
No transcript available for this episode.
