# Data Cloud Keynote Deliver Customer Success Through Data Auto-transcribed by https://aliceapp.ai on Tuesday, 17 Sep 2024. Synced media and text playback available on this page: https://aliceapp.ai/recordings/YAoT9VL64xqtsM_wNu0d3P78l2jeOGCc * Words : 8,126 * Duration : 00:51:00 * Recorded on : Unknown date * Uploaded on : 2024-09-17 23:23:10 UTC * At : Unknown location * Using : Uploaded to aliceapp.ai ## Speakers: * Speaker A - 14.87% * Speaker B - 25.2% * Speaker C - 23.58% * Speaker D - 19.3% * Speaker E - 17.06% ---------------------------- Speaker A [00:00:13] For us, it was a natural notion to go to data cloud. Speaker B [00:00:15] It is the center of the universe. Speaker A [00:00:17] In that everything else hangs off of. Speaker C [00:00:19] The data about our customers. Speaker A [00:00:21] And it seems like a simple thing like turning data into insights is super important to developing that 360 degree view for the purpose of serving that customer better. Datacloud enables us to bring all of. Speaker C [00:00:36] Our uh, different digital streams together with our CRM data. Got your website, your configurator, uh, vehicle, data marketing, uh, cloud, uh, autocloud and commerce cloud in order to see absolutely everything that a customer is interacting with, all under a single pane of glass. Speaker D [00:00:52] Datacloud has enabled us to centralise all of our customer data and expedite the customer journey throughout the purchase cycle. It's allowed us to do cross sell and upsell, uh, opportunities. And so that's been a really great sales revenue driver for us. Speaker E [00:01:10] When it comes to data cloud and. Speaker C [00:01:12] AI for our sales team, this is a game changer. Speaker E [00:01:15] To see a system that was going to unify our data, uh, gaining better. Speaker D [00:01:18] Insights for our sales team while avoiding any kind of swivel chair super charges. Speaker E [00:01:23] Our business. Speaker C [00:01:27] With data cloud, we have been able to put together all of our salesforce instances, bring that together to deliver those customer experiences that uh, make their businesses better and all their lives better. Wouldn't have been able to do it without data cloud. Speaker D [00:01:45] Please welcome EVP and GM Unified data services in Einstein Salesforce, Rahul Aradkar. Speaker C [00:01:54] Wow, what a great video. Thank you, thank you. What a great video. I can see that video over and over again. It's inspiring and quite frankly, it's emotional for me. Some of the best brands, some of the best companies in the world are depicted in that video. There are hundreds more successes. Some of you, some of the leaders from these companies are here, Tony and a whole bunch of other people who are watching online. Thank you. We're glad you're here. This is the first dedicated data cloud keynote ever. It's exciting. So over the last three years and eleven months, I'm counting every month, we have seen unprecedented success for data cloud. You have made data cloud the fastest growing organic innovation in the history of our company. I would argue it's one of the fastest growing organic innovations in the history of enterprise software. Now, all of that is possible because of ongoing collaboration and feedback from you. So as the GM of data cloud, from the bottom of my heart, thank you. So all of you, what we do with you, all of you are the heart of everything that we do with you. Now, you are at, ah, the forefront of the future of work. We've been talking about AI, but we know the value of AI is when humans work with AIh in collaboration. Now, how do humans and AI collaborate together? One of the key ingredients it hinges on is reliable, trusted, secure, harmonized and unified data, whether it is structured data, or whether it's unstructured data. Now we are in the third generation, third evolution of AI. No matter where you are in your AI journey, the importance of data continues to increase. Whether you're an automotive company that's driving propensity to buy predictions using a whole bunch of engagement data, structured and unstructured data, or whether you are a retailer generating content pertinent to your brand voice. No matter where you are, the importance of data continues to increase. Now that we are in the third evolution of AI, which is assistive agents, you heard about it this morning, agents in agent force. You can picture an agent powered by AI working with a human, resolving a customer service incident in real time with data that is trusted, governed, harmonized, unified, coming from multiple different systems, and doing that in real time. Now, you can imagine the complexity of that, but that is what's happening in terms of importance of AIH as we go up the evolution of AI. But businesses have to deal with challenges, and they range from the simple ones that I'm showing you on the screen. Sounds simple, but getting a harmonized view and a unified view of a customer is a challenge that we've been talking and we've been theorizing for the last few decades, at least last few decades that I know of. It is a non trivial task. That dream that we've been talking about is the reality. Now, just a few weeks back, we were having a discussion with the world's best entertainment brand, and they told us that they have 150 data sources that talk to their customers, talk about their customers. They're all fragmented and siloed. They sit in lakes and warehouses, and for good reason. They're sitting there for lakes and warehouses for different functions and utilities. They asked us if we can help them. The answer is clearly yes. That is exactly what we do. You can imagine you're enjoying the experiences of this company's online offerings, as well as offerings that they have, physical offerings that they have in their parks. And you can imagine a concierge experience where you have a real time automated actions and automations that you're looking at. And this following you everywhere you go from a vip experiences, that is the proverbial guest 360 and that is fluidity of data. Now, to get those challenges solved, customers have to solve the non trivial or the proverbial chasm that you're looking at on the screen here. So how do we do that? That is the reason why we built data cloud. Data cloud brings in all your data, unifies that and provides you a uh, unified profile of every customer. And data cloud is built on Salesforce's metadata framework. It is deeply integrated into the Salesforce platform. What this means is that trailblazers and app developers now can use all the richness of data that's in data cloud in concert with all the tools that they're used to and come to love. Things like flow for automation, Apex lightning reports and dashboards. They can now use that seamlessly. With the richness of data cloud. It is the evolution, it's the next evolution of the Salesforce platform where data is at the center of everything that we do. Now. If you take a look at the overall platform, you saw this, Mark was talking about it in the morning. Data cloud is the heartbeat of the Salesforce platform. And the Salesforce platform allows agents powered by AI working alongside humans, driving actions, automations and actionable insights, all in pursuit of delivering customer success. And perhaps the most important thing here is we're doing all of this with governed, secure, reliable, trusted, harmonized, unified data. Now there are many customers who have really seen successes and transform their companies and we are honored to have partnered with a few of them. We have FedEx here on the screen and we have Heathrow here on the screen here. I was having uh, breakfast with uh, Tony, who is on the screen here this morning, and he said a very interesting thing. He said, shopping is shipping. So through this entire network they have multiple petabytes of data. As a matter of fact, they generate a petabyte of data a day. And what they do is bring in all of the data that is sitting in lakes and warehouses, which they're using it for different reasons, for doing their shipping optimizations, for their network optimizations. They bring the data into data cloud in a seamless manner, unify it with their customer engagement systems, in this case their CRM systems, and drive seamless experiences across sales and marketing and they're extending that out to customer service. And when I was talking to him, what really excited me most this morning was they're extending it out to shipping logistics. And the examples he used were pretty phenomenal. I'll let him talk about this. He's going to be on stage very soon. And then you have Heathrow. Heathrow has a whole bunch of airline information, schedule information, passenger information, et cetera, sitting in lakes and warehouses just like FedEx. Now, Heathrow wants to provide delightful end to end kiosk to kiosk experiences that they do by bringing information and data that is sitting across all these lakes and warehouses and using customer engagement systems that are built on Salesforce and Salesforce platform. These are just two examples. There are hundreds more examples who have seen this success. I refer to the fastest growing organic innovation in the history of Salesforce. Now, while it's a pride that we have, we are humbled by it, but these are the kind of successes that are really driving the usage and adoption that you saw on the screen. Now, I want to shift gears and talk about how these customers are achieving this success across these three chapters, if you may. How do you create a profile, drive action from it? And how does it ground AI with power trusted data? We have three chapters. We have three fantastic presenters here. I want to get started and invite Tarana Whitefield to the stage. Speaker E [00:11:12] Hello everyone. Welcome to Dreamforce. It is so good to be here with you today. Now we're experiencing a massive shift. AI is revolutionizing how we work and how we engage with our customers. But we know that AI success hinges on having high quality, relevant data. So let's go ahead and dive in to see how you can create your customer 360 and unlock the full potential of AI. Now, nearly every company that we talk to has some form of data strategy. In our customer conversations with all of you, we hear that one of the biggest challenges is connecting all of your disconnected data and putting that data to work. Now this includes your structured data across apps, uh, APIs, data lakes and data warehouses, along with your unstructured data PDF's, audio and video files. By analyzing unstructured data, you can uncover deeper patterns and sentiments that structured data alone might miss, leading to better personalization and better decision making. You need to be able to integrate and harmonize all of this diverse data to truly create that unified customer view. But where do you start? Well, with data cloud you can connect and unify any data from anywhere. Now we're so excited to announce that we have over 200 out of the box data cloud connectors and pre built APIs from Mulesoft ready for you to use. So now you can bring in all that data into data cloud without having to write a single line of code. But what about your Salesforce data? Well, we have a powerful CRM connector that allows you to connect to any of your CRM apps across sales, service, marketing, commerce. You can bring that data in the data cloud in near real time. But that's not all. We're giving you flexibility when it comes to connectivity. We're introducing zero copy integration and our new zero copy partner network. So now you can virtualize and act on all of your data from your data lakes without having to physically move that data. And with file federation, you can securely access and analyze large data sets from your data lakes, ensuring that you're staying compliant and boosting your performance. Now, one of our latest enhancements is data cloud one. Using zero copy technology, you can implement a central data cloud across multiple salesforce, orgs, accelerating your unified customer view without ever having to write a single line of code. Now okay, so how many of you in here have sent a, uh, text message or a photo from Dreamforce today? See the hands? All right, I think everyone I know, I have too. Well, today we're connecting through text, video and audio. That's the unstructured data we want enterprises to be able to capture and put to work. Introducing data cloud vector database. So now you can extract rich insights from your PDF's, your video files and your audio files. So you can retrieve information from your customer service phone calls, your confirmation delivery photos and your sales agreement PDF's. Now you can take all those insights, put it back into your CRM and power agents, analytics, automation and your customer 360. Now that's the power of putting unstructured data to work. But what about data security and governance? We know this is top of mind for every single person in this room, and different enterprises have unique security needs. But it's up to all of you admins into a room to ensure that the right teams have access to the right data at the right time. Now we know you're dealing with hundreds and thousands and columns and tables of data. So managing security and governance, it's hard. Now, data cloud was built on our Salesforce platform, so you know it's secure and trusted. But we're giving you extensive security and governance features built natively into data cloud. Introducing policy based governance and AI based tagging. Now the way this works is you first identify the tags that you want on those columns and rows and then you define those policies. So for example, I can give tags such as confidential data, security data, or financial data, and I want to give Rahul only visibility to financial data. It's as simple as that. And we also heard all of your feedback and we listen. We're also introducing private connect so you can securely connect to all of your data warehouses. And using customer managed keys, you can rest assured knowing that your data is secure while at rest. All right, so I know you want to see some of this in action, so why don't we go ahead and jump into a demo. Great. So freight is a leading shipping and logistics company and they're using the power of data cloud to create their customer 360 and freight employees, well, they need a comprehensive view of their customer across sales, service, marketing and every touch point in between. This includes past orders, shipping history and all that rich engagement coming from the web. Now this standard contact card does not fulfill those needs. Let's go ahead and see how we can enhance this view using data cloud. Let's start by pulling data into data cloud. Now for all of your Salesforce data freight has uh, access to our CRM, um connector. They can go ahead and use that connector to connect to any of their CRM applications across sales, service, marketing and they can bring that data in using no ETL. Now for external data, whether structured or unstructured, freight has access to over 200 out of the box data cloud connectors. Now using zero copy technology, freight can seamlessly connect to Google, Bigquery, databricks and Snowflake without having to physically move that data. Now this process ensures access to real time data and consistency across the this data is then mapped to Salesforce objects so it can be used across the now Salesforce's Unified Metadata foundation maps this data to data model objects so you can create that harmonized view of Damiana. Now freight also uses identity resolution rules to match and consolidate customer identities across various touch points. Now this comprehensive view is crucial for customer satisfaction, personalized marketing and trust because now Damian shows up as one person and not 50. Now once all that data is mapped in, the Salesforce offers flexibility to build, bring your own model or use any of our out of the box AI models. Now freight uses model builder to create their own custom AI models specific to their own business needs, such as predicting customer churn. Now models alone aren't always helpful. Your users want to know what they should be doing next. Now the good news is model builder models are now actionable. So here I can actually select the variables that are going to power those recommendations to those users. Now these recommendations are going to lead to more impactful and successful outcomes and better customer experiences. Now freight also has access to AWS Sagemaker to import advanced predictive models. That's going to help them extend their existing data foundation. All right, so let's pause for a second. Remember this customer profile that we saw a minute ago? Well, this is fine for basic customer needs. But let's go ahead and use all that data and predictions and enhance this view. Bam, there it is. Here is a comprehensive view of Damien. Now you can see all the rich engagement coming in. You can see even a tableau metric built right in all the shipping history. Now freight has completely transformed data management, delivering a true customer 360 using zero copy integration and predictive insights. And that's how you can connect, harmonize and unify your data to create your customer 360. We're just getting started. Now that you have all this data unified, we're going to show you how you can put that data to work and to walk us through the next chapter, please welcome to the stage Desiree Motamendi. Speaker D [00:20:27] Thanks Tarana and welcome everybody. This is my first Dreamforce and I'm so excited to be here with all of you today. To deliver the right customer experiences at the right time, teams need access to the right data at the right time. Most organizations have a data strategy in place, but the reality is that only a few, usually the it organization, who have the access and the ability to do anything with that data. As, uh, Toronto shared data cloud is empowering you to unlock the true value of your data with a unified customer profile. Instead of sellers following up on the same sales leads, sales teams can now identify upsell and cross sell opportunities. Instead of poor customer service service teams can now proactively resolve customer issues by anticipating their needs based on past interactions and behaviors. Instead of generic campaigns, marketers can use data to personalize all their campaigns, personalize all their campaigns so that they can tailor to each company's interests and preferences. With data cloud, Salesforce is democratizing data. Data Cloud is the center of the Salesforce platform. It really helps enable, oh gosh, I'm so sorry guys. Salesforce is, yes. Thank you, thank you, thank you, thank you. Data Cloud is the center of our Salesforce platform. It helps enable processes like data triggered workflows. It also helps, uh, perform in depth analytics with predictive and, um, calculated insights. And it also helps build AI apps and agents grounded in reliable data. With our low code and pro code tools, anyone within your organization has the ability to act and act on that data, which enables self service and empowering teams to make informed decisions and drive innovation. Data cloud is truly helping our customers build the customer 360 and we have seen some customers who have seen great success. Wyndham has seen a 55% increase in franchisee resolve customer care cases. They're using data cloud with mulesofthe to enable seamless data fluidity. Data fluidity. Sorry, apologies again guys. Data fluidity. And this has helped them. Oh yes, data fluidity. This has now helped onboard new, new uh, hotel owners. Being able to now bridge the gap between hotels and call centers. Fisher and Paykel has seen tremendous success. They're a premium appliance brand company where they're now centralizing and enriching sales data in sales cloud with the power of data cloud. And now they're enabling reps to spend less time on gathering information and more time on productive conversations. This has helped them improve by 3300 hours per month of improved, um, automation and self service. And Muskoma bank has seen a 45 minutes cut in manual data per new loan. With data cloud, they're helping unlock trapped data across 66 different legacy systems. Oh yeah. And with subsecond real time and Einstein personalization powered by data cloud. I apologize guys, I'm so sorry. And ah, with subsecond real time and time personalization that's powered by data cloud, teams have access to data faster than ever using both structured and unstructured data. Now sales teams can get access to company, uh, purchase history and personalization in real time, allowing for personalized recommendations. Service teams can access to up to date interaction logs enabling. Oh my gosh, guys, I don't know what it is. I was doing this all earlier today and I was doing just great and then I get in here and it all just goes all the way away. So I apologize guys. So service teams have access to up to date interaction logs where they can now resolve customer issues quickly and marketing teams can use the same data to now be able to create target customer segments for more promotional campaigns in real time. This ensures this cohesion, increases sales, improves customer satisfaction and delivers more efficient marketing campaigns. But what better way to show you than with a demo? We're going to bring freight back up on screen and we're going to show how they're bringing their unified data to work across marketing and sales. The marketing team can create a new custom segment for corporate accounts with high engagement. The team uses Einstein to create the new segment and the team uses uh, Einstein to create the new segment. Oh my gosh. The team uses Einstein to create the new segment and they can include additional related attributes with the AI suggestions, the marketing team can choose what to include and close out the segment with data cloud, freight can use uh, ad preferences can choose advertising partners like Amazon, which now includes ad insights that's only unique to Salesforce. No one else has this functionality. They can also use advertising partners like Meta and Google Frake can use insights like affinities and demographics to enhance their ad campaigns. This ensures that the most relevant ads is targeted to the right customers so you can get more out of those clicks. And as a marketer, I know how important those clicks are. Analyzing the performance of those segments is top of mind and drives ROI and decision making. At freight, the marketing team can track effectiveness of journeys, segments and campaign engagement. But the journey doesn't just end here, it also continues on the sales side. Freight can create automated triggers that can initiate action. This data triggered flow creates a slack channel for corporate accounts like Damien's, who see interest in international rates on their website. This automatically notifies the sales team so they can swarm and offer assistance back. On the 360 view of Damian's account, you can see that April can choose from a set of pre configured prompts that has taken data from data cloud to populate the email. The sales rep can then now create a drafted, personalized, AI generated email grounded in data with real time engagement. Datacloud is working hard behind the scenes to be able to power all your customer experiences and is now ready to tackle the next generation of people and AI. One trailblazer doing this today is FedEx. They have seen tremendous results with their shipping volumes using data cloud. Let's go ahead and hear from them and roll the film. Speaker A [00:28:41] Over the last 50 years, FedEx has built one of the most powerful and unrivaled physical supply chain networks in the world. We move 15 million packages a day across 220 countries that we serve. Speaker C [00:28:55] When you add that up on a global basis, annually, we're facilitating 2% of the world's gdp. Speaker D [00:29:03] That's $2 trillion a year of information. Speaker C [00:29:06] That is attached to all of those packages. Speaker A [00:29:09] From that physical network, we create one petabyte of data, uh, every single day. So how do we take advantage of that data and turn it into something that makes your business better? Speaker D [00:29:21] At FedEx, the customer is at the center of everything. And before data cloud, we definitely had siloed data. Speaker E [00:29:29] So customers may be on.com one day. Speaker D [00:29:31] They may have 19 shipments in flight. And when they come to FedEx, they expect us to know what they've been doing. Speaker B [00:29:39] So sales has a view of the customer. Marketing may have, uh, the same customer, but they're looking at it from a different angle. Then you talk about shipment. So it takes time to collect that data, most of the time over three weeks. With data cloud, what we have now is real time. Speaker C [00:29:55] When we think about data cloud and zero copy, it was vitally important to us. We have all of our operational data and pricing data and costing data sitting inside of Azure. So we were immediately able to connect without having to copy all of our data. Speaker B [00:30:11] We started with a million records and we narrowed it down to 12,000. Speaker C [00:30:15] And that in turn is the data. Speaker B [00:30:17] That'S activated into marketing cloud. Speaker E [00:30:20] The ability to connect our marketing cloud. Speaker D [00:30:23] Our.Com channels with our sales instances and serving it up to Salesforce so that we can take action on it is what data cloud delivers. Speaker B [00:30:33] As the journey goes on, all the information is recorded and then ends up back into data cloud. So now we have learned something about the cluster. Speaker E [00:30:41] We can shift all of those. Speaker C [00:30:42] If a customer is having difficulty online, we can direct it to a live customer service agent and the salesperson can reach out or we can send them an email. So we have brought all this together through datacloud. Speaker A [00:30:56] We love powering experiences that allows our brands and our customers to grow their business. Delivery predictions on the FDX platform is just one example of ways that we're helping customers increase their revenue and win together. Speaker C [00:31:07] I think we're just really at the beginning point of generative AI and all the capabilities that it's going to be able to do to augment the human interaction. AI will help humans become more human. The culture at Salesforce of being customer obsessed is a leading attribute. The sales force is constantly innovating and constantly creating new value. I just can't say enough about how well the teams have come together to delight our customers. Can't ask for more than that. Speaker D [00:31:43] Joining us on stage is Tony Krieger from FedEx. Welcome, Tony. So it's pretty amazing because in that video you talk about one petabyte of data per day that you process at FedEx. Tell us what are the technologies and strategies that you use to be able to take advantage of that data? Speaker A [00:32:02] Yeah, thank you. It's really great to be here. Great to see everybody. I'll just start by saying at FedEx we're on a mission to make supply chain smarter for everyone. And as we think about that, there's no way you can transform your business like that if you're not leveraging data. About four years ago we started our journey, um, we just kind of mentioned data all over the place. A petabyte of data, um, every single day. And in the video you heard that it was not only the size and scale of the data, it's the richness of the data. So if you think about it, $2 trillion a year of GMV, we move around the world every single day you take that with over 700 aircraft, over 200,000 vehicles, over 500,000 fantastic worldwide employees, and you bring all of that together and you can create some pretty valuable things. We're thinking about it in three different ways. One, we're leveraging this data to fundamentally change the way we run our network, right. We want to continue to be the best, most powerful transportation network so that we can help deliver the promises that our customers give to their end recipients. The second piece is we're digitizing supply chains. With our customers, you can't manage what you can't see and uh, what you can't measure. And so we're helping digitize our customer supply chains. And then the third, um, that we're very excited about, um, and that we're really um, starting to gain some momentum on is we're actually moving up the e commerce value chain and we're doing that through this past Sunday. We actually uh, made generally available our FDX platform and we uh, private previewed that starting in January. And uh, let me bring it to life for you. If you think about an e commerce company, none of us shop anymore without the expectation of understanding at purchase time exactly when it's going to be there, right? So in the pre cart, uh, we're now powering who better than us, the transportation company and who knows a lot about it, to actually uh, help customers predict that in cards we're helping, say order within the next 1 hour and 23 minutes is guaranteed there Thursday between 10:00 a.m. and noon. And we're layering our insights. This is the importance of bringing our network data and our customer data and our service data and our sales data and the ability to bring it together, which is obviously what the data cloud is allowing us to do. We then move that to the post purchase where we help power the post purchase experience. And that's where our sales teams engagement, our customer service engagement. When you talk about data all over the place, it's really hard to create a meaningful experience and certainly create products that helps our customers and uh, our brands and our retailers actually grow their business. We want to be, and we are being more than just the last mile transportation provider. Uh, we're partnering with them m in helping them grow their business. Speaker D [00:34:31] That's awesome. So we've been working together for a while now, Salesforce and FedEx. But I'd love to hear what this partnership means for your business. But more importantly, what are you looking forward to in this collaboration in the future? Speaker A [00:34:41] Yeah, it means quite a lot. I mean we've been working together for 15 years now. Um, and that partnership has allowed us to think through the, um, way that we build out both the service side, our customer service side, and obviously our sales side as well. Um, and so as you look at the last year, as we've really leaned into data cloud and we were talking about it just this morning, um, just to show you size and scale of how we're using data cloud. In the month of August, we had 235 billion rows of data in data cloud. Speaker D [00:35:11] A lot. So if anybody's curious, that is a lot. Speaker A [00:35:14] Yeah, if anybody's curious if it can scale, um, we have pressure tested it for all of you. Um, but again, and I just want to reiterate, like, a lot of data is not that useful if you can't curate it and you can't generate insights that then you can take action on. And that's the journey that we've been on, um, over the last year. We take that data that lives in Azure, um, and without having to copy it. And this was a big request that we obviously had, is I don't want to move that size and scale of data all over the place in order to make it contextually relevant to create that insight and to drive that action. And so we are literally leveraging data cloud connectors and we're taking advantage of all that operational data, all that customer data, all that sales data, of course, that lived in different ecosystems. And you bring in of course, our own proprietary enterprise data platform because that much data, um, we want to manage it, we want to own it. Um, but you've got the power of Salesforce and the power of the salesforce cloud and so we don't have to take it and move it all around. We connect those together and the value that we've had just in one year is pretty phenomenal as we think about recognizing potential attrition, if you think about it, um, a weather event that potentially causes a package to have a delay on it, um, we need to know that. Our sales team needs to know that. Our customer service team needs to know that. So we take the operations data, we match it with our customer and our sales data and the interactions and the relationships that we continue to have with our customers, um, are truly fantastic. And so we really appreciate the partnership. Uh, we look forward to growing significantly in this space. I'll tell you, like, we're in the process now of completely reinventing. And you said kind of what's next? I think taking this and now moving it into, um, how we leverage AI in our journey, orchestration is really where we're headed. Right. And I think being smarter, um, being obviously more efficient, and then I think all of us as consumers and all of us as. As companies ourselves, want that interaction to be, uh, pretty seamless as well. And together we're doing that. Speaker D [00:37:05] That's amazing. The power of scale, guys. Um, thank you, Tony. But before you go, before you leave the stage, I want to be able to introduce our new data community, datablazers. And because you've been part of this data journey with us, we'd love to give you our data blazer sweatshirt. So now you can wear this and be an official member, which is really awesome. And we have some data blazers here in the audience as well, so thank you. Appreciate it. Speaker A [00:37:30] Thank you so much. I'll probably wear it. Thank you. Speaker D [00:37:32] And for everyone else in the room, our datablazer community is now live. This is for it, folks. Business developers, data practitioners, anyone who wants to unlock the true value of their data by joining, you become part of an elite group of data ambassadors. And now also here at Dreamforce, if you become a member, you have the choice to actually get a datablazer pin set. Um, hopefully you can get some because I hear they're running out. Um, but really, this will be part of a unique journey and a symbol of your data journey with us. So sign up today, and I hope I see everyone in this room with a data blazer sweatshirt next year. All right, so you've heard from Tarana about how you can unify your data, and I tried to talk about actioning your data, so I apologize for all the mess ups, but you're probably wondering how you can connect data and AI together. In this next chapter, we're going to talk about how you can ground data with your AI apps, and of course with Agentforce, because you've been hearing all about it. So please welcome to the stage Mk. Hi. Speaker C [00:38:30] Awesome, Gus. Speaker B [00:38:32] Let's give it to Des and Tony. All right, good afternoon, everyone. I was joking that I'm a walking ad for our thing. It's data plus. Agent plus human is customer success, right? Your success is our success, so that's kind of our mantra. All right, speaking of AI and agents, you saw Rahul present this before, right? We've been through a predictive, generative, and now we're in the assistive third sort of wave. Now, as these waves go on, data becomes even more important. Till probably last week, I think if you asked chat chip, how many hours since Strawberry said two, right? There was a joke around it, and we all just joke and just forget about it. But imagine if that assistive agent is going to tell you your next quarter result or next quarter, what the growth is going to be or what campaign you need to do. You've got to trust that data. And same thing. Now, when you go to the next level with autonomous agents and you're going to put that agent in front of your customer, the stakes are much, much higher because it's your brand reputation at stake. And it's also important to know that over 80% of the data in the world not structured, it's unstructured. That's in your email, in your case, incident reports, in your documents and words and so on. Now, your service agent might be able to look up five different websites, read five different documents, and answer some question. But imagine if somebody is asking your agent one single line that says, hey, what should I do about this? That agent better have access to all that data, right? So that you can actually answer and give the right answer. That is why we built unstructured data right into data club. So just like you have all the structured connectors, you can do the same thing with unstructured data. Now, this unstructured data could be columns and rows in your case object, your knowledge articles, or it could be PDF's or sharepoint documents that are sitting anywhere. Again, either zero copy or through ingestion, you can bring all this data. And then we do some interesting things. It's not just we just copy it and leave it as it is. With our atlas technology, we have some built in chunking techniques. So chunking means you take this large document, it could be gigabytes, terabytes, and we want to chunk it into smaller paragraphs. And you can use different chunking techniques, like paragraphs, or in the case of video or audio, it could be the conversation boundary, or windows or sentence and so on. And then we apply a bunch of enrichment to it. And this is why we can make that unstructured data work. And with this enrichment, we can add some data Q and a, some entity extraction, all that stuff. But more interestingly, we can also create and use different embedding models. We have some of the world's leading researchers in Salesforce, and with their SFR embedding, which is one of the hugging face uh, top models, we actually then create an embedding. Embedding is nothing more than a mathematical representation of all that chunk or that unstructured data. And because it's a mathematical representation, you can also work across languages, you can have a french document or an english document, it'll probably map to the same thing if the content is semantically the same. Now after we've done that, you've created the index, that is the vector database. You've got to be able to retrieve it. So we have a whole bunch of retrievers that we have created that includes all the semantic search retrievers that can actually look it up by semantic. We have data graphs, those are powerful real time sub 2nd 360 kind of graphs that you can get, or hybrid search that includes keyword and semantic search. Now to that you can use these simple retrievers in from studio that you earlier saw Sanjana demonstrate in the main keynote. And with that the prompt then grounded with the LLMs. It could be our LLMs that we use out of the box or your own LLMs as well. This retrieve augment generate that is rag and that works across structured or unstructured data. Now with this, basically you can answer any question. Now the first thing you're going to ask is can it answer any question, what about all my secure data? And that's why we had that security. Remember Tarana talked about, we are introducing this new AI based classification, uh, tagging and policy based access control that also works on your unstructured data. In fact, we want to go one level up in the next releases where we're even going to do at a chunk level at an unstructured text, we can say, oh, this text should be secure. The next sentence is probably, ok, so we are really elevating this kind of policy based governance to the next level with all this. Now you can expose your AI to your customers, you know you can trust, you know it's grounded in the right information, structured, unstructured, you know it's secure, and so you can put your brand reputation in front of your customers. Speaking of which, that other thing that you see here on the right side, if you can recognize that, that is the Askastro app. If you have downloaded Salesforce event app, you ask Astro, it says ask Astro beta. That is actually powered by agent force. Uh, and all this unstructured data, it's grounded in data cloud. In fact, it was a journey by itself. It'll write a blog on it. It used to be a diY. And then we within three days moved it all into this whole platform. And that is a testament to what's possible. Now, enough said with all this stuff, uh, let's go hit the demo. All right. April and Adam right there at the demo booth. Now, one thing I'm going to tell you here, you're probably seeing a lot of demos. You're probably all thinking, you know what, some figma. These guys have just cooked something up. They just keep clicking around. I can guarantee you every one of this is production. It's running on my production. I know every click works. Uh, here. So, all right, so let's come back to Damien here, right? So Damien has a shipment that, uh, like Tony says, like many other, uh, people will be asking, hey, where's my shipment? Right? And interesting thing with daemon, he doesn't like to talk to people, so he's typing right there. And you saw the answer came right there. But. But if you followed what Tarana said earlier, that Damien, if you actually flipped the, uh, thing, the Damien is actually a unified profile, and his shipment information was sitting in Snowflake. So how did we bring this over to the agent? Now, that shipping agent is the mystery. In the shipping agent, we have created two topics. One topic is for shipment tracking. Now, what that topic, if you click on the topic, it basically gives up set of guidelines to say, this is all about shipment information. And the action we are telling it, uh, is to say, go look up the shipment details from Snowflake. And so that's basically what it's doing when you say, where is my shipment? Now, if you want to really see something interesting, April, you want to try this. All right, she's going to go. This is the backend, Snowflake, that's powering it. Uh, we're going to go update that statement right there in Snowflake just to show you how real it is. Then we're going to go back to here, and then let's ask, where is my shipment again? Now, if everything works, like I said, it's all live. If everything works, uh, and just for fun, yesterday morning, it wasn't working because OpenAI was down for, uh, as you can see here. Look here. That was a real time update. We updated snowflake through zero copy. It came, it was there in data cloud. And when Damian asked the question that, uh, cookie id device id was mapped to, actually, Damian has 46 different profiles. It got mapped, and we then correlated to this and pulled that information and showed it. But that's structured data. What about unstructured? Now, Damian might have a question like, hey, I want to change the schedule. What do you want to do? This information is not sitting in any column or row or anything else. It's sitting some PDF document somewhere. Now, this agent is going to actually answer that question, too. Uh, there you go. Oh, sorry. See, it's real. It'll answer. Now, if you go back to how it's actually working, you can put that and maybe go to the vector index. Behind what's going on here is when she asked that question, the topic we automatically figured out that was about knowledge FAQ. So that was grounded in data Cloud's search index. So here you see a bunch of search indexes created on shipping FAQ, wine FAQ. I guess somebody's a wine lover. There you see a lot of interesting indexes created. Is as simple as clicking new and then choosing the vector index you want to create. You can do an easy setup or an advanced setup. With the advanced setup, you can choose all your chunking strategy, your embedding strategy, and everything else with easy setup. Um, one click. It just works. Now, once that is done, this is that prompt studio that you can actually use to then ground it. Now here, as you can see, we have said you need to act like a customer service agent. Put all the guardrails if you want. And then also you can see, I, uh, put the case description in there and also saying, retrieve knowledge for that case. So then you can actually then ground based on the question that is coming, it will go look up that index and go answer that particular, uh, answer. So that was the whole, as you can see here, this would be the actual example here. As you can see, the agent figured out the right action. It picked the general FAQ, it did the actual index lookup, and then returning the answer. That's how we're bringing structured and unstructured together. But you might ask, okay, is it all just AI? Is it like, can we just use it here? No, we built a platform. What it also means is that all that structured and unstructured data is available via SQL interface. So what can we do with that tableau? Now you can actually explore and analyze all your unstructured data in tableau. You can ask unstructured question, look at that. What customers are talking about lost shipments. And Tableau is able to do a semantic comparison. You don't have to say lost shipments in any of your description. As long as it's related to loss shipments, it'll find it. We have cross correlated with the case object, and then we have cross correlating with the contact that you see here and then showing that geolocation. And just to prove you can see, you can click on it and you can actually take you back to that service cloud instance as well. So this is how we're bringing data. Uh, agents, AI, tableau, CRM, altogether by the power of data clean. All right, back to you. Speaker C [00:49:06] Ok, that was amazing. Talk about live. You saw some hiccups there, the fun of live code. That's amazing. So give it up to Tarana. Des is first keynote, first data cloud keynote on uh, salesforce moodily and of course April and Adam. So thank you. So we have seen, we have seen quite a few successes here for data cloud. What you're seeing on the screen is a few more successes. You're at the home stretch, you've got a minute more. I'm going to wrap it up. A uh, few more successes here for customers here. So data cloud has been on a tear. We have been innovating and we ship every month. And what you're looking at on the screen, pick up your phones, take a picture of this. These are capabilities that we shipped in the last couple months and what is coming in the next few months. Data cloud, the innovation train continues and it delivers based on your feedback. Now uh, just to wrap it up, the fun just is beginning here. There are multiple data cloud sessions there, the roadmap sessions. There is a trailblazer forest where you can get your hands on, you can work with data cloud and with agents, create your own agents and of course you can get your pins that uh, Des mentioned as well. I want to leave you with this. DES talked about the datablazer community. This QR code will allow you to join that datablazer community, the like minded professionals here. We're just getting started, we're just getting rolling with the datablazer community. Um, amazing content already on the community. So with all of that, thank you and please have a great dreamforce and don't forget to take the survey here and give us your feedback on the social media channels as well. Thank you all.