Salesforce, Inc. (CRM) Earnings Call Transcript & Summary

Hi, guys. This is Amin from Salesforce team. I'm part of Platform Solution Engineering team. And before anything, I'd just like to say a warm thank you to everyone. You might be Salesforce customers or partners or Salesforce employees. Thanks for joining in and give me an opportunity to ask -- for talking about connected data driven apps, right? So without further ado, a quick forward-looking statement as usual. So we might be discussing certain features and functionalities of Salesforce products, which are currently in road map, right? So please make all your purchasing decision considering the current available features, right? Having said that, let's get started. Now when we talk about digital, digital is no longer just a buzzword. There are a lot of improvements, I would say, a lot of adoption that has happened over the period of time. Probably 5 years back, digital just is being used around and thrown around like anything else. But now it's -- you can see that shaping a lot, especially in the India market. When we talk about mobile first customers in the world, we see that there is 20x more engagement that is happening through mobile apps. We see that the push notifications are going more than 300% cap rates are there. And more and more orders are processed through digital channels now. And especially when we talk about India, in India region, we see that now, there will be 840 million users by 2023. Not only that, smartphones are like in the hands of every individual, right? And there are 700 million smartphone users by 2023. And having said that, digital, especially digital commerce is going to be adopted like anything. We expect that we will have a 3x more share in digital market, retail market. And India will be the fifth in consumer market by 2025. Now all this is happening through the -- a lot of initiatives that has happened in India, a lot of initiatives that happened across the world, especially in terms of broadband Internet, 5G, which is being in use and of course, the smartphones. But delivering that consumer experience or customer experience, when you look on a digital customer experience, is often met with various challenges, right? So you will see that it takes a lot of time and effort to build a mobile app and make it available to your customers and partners. And it takes a lot of effort in terms of building those experiences, testing them at scale and then deploying it on a highly scalable environment. It takes a lot of effort and energy, right? Now what we can do to simplify that. So when we talk about a consumer experience or a customer experience, typically, these are various kind of features and functionalities that you would like to offer. When you have a connected app, you would like to target the digital advertising that could be through a web portal or a mobile app as well. You might want to target them as a social media. You would have an e-commerce platform where you're selling your services and offerings. You would like to send them personalized offerings and personalized recommendations as well, often with the power of AI. You would like to have an effective order or service management, and then you'll lead a good amount of inventory in warehouse. So this wheel that you see right now, which goes across the consumer service kind of covers the overall journey of our customer, right, from the awareness, acquisition to conversion and finally, for retaining the customer and increasing the brand loyalty, right? And to deliver such kind of an experience, there are a lot of building blocks that needs to be delivered. So let's see what would take to deliver such kind of a platform, right? When we talk about such kind of a platform, this is a typical, I would say, a functional architecture of building block diagram of a customer app or any app for that matter. You would have a system of automation, which will take care of your of kind of auto-scaling, deployment, DevOps. You would need security at the forefront of it. You will make sure that your customer data is protected, all the sensitive details protected. And then you will have key layers. The key layers would be system of record where all the data is stored, and this could be spread across multiple, I would say, systems. You could have system of record in a CRM system where you maintain the customer profile. You might have SAP or ERP systems where you have your orders and inventory managed. You might have e-commerce platforms where you have your cost management and all those journeys for our experience, this will come out of experience. But typically, these layers will be -- one will be a system of record. Then on top of that, you have a business logic. Often you want to have a containerized business logic, so you have critical level of scalability. So that's where all your workflows and business processing sets, all your logical, reasoning, rules, all these things goes into business logic. And then you will have a presentation layer, which is maybe in the form of web or a mobile app or maybe other various channels. Now all of this what we call is a system of innovation because this is where all your business logic, customer journey mapping, all those key criteria, which will help you deliver a premium customer experience sets and designs. Apart from that, you might also have a system of interaction where you want to expose all these functionalities as a REST API, for example, or through an event-driven architecture. So not only customers can interact, but also other various other applications can interact with your business logic. For example, you might want to expose APIs so that customers can place orders through a smartwatch or through an IVR or through a bot. It will take a system of interaction to expose the end points so that those transactions, those journeys can be fulfilled across channels, right? And then, of course, nothing -- not any system in the world today can be completed without a system of insights. Now the system of insight is basically where you will generate KPIs, where you generate your recommendations and personalizations, right? So this is the system which not only will be for the reporting and analytics purpose, but it will give you advanced analytics as well, which includes your ML-based algorithm churning out, propensity to buy our best recommended product, a mixed distraction for you, right? And then there could be a lot of other integrating systems that you will be interacting with. For example, if you take an example of a loan origination system or NBFC. In our loan origination system, there will be a lot of third-party system, including in terms of rating systems or credit rating systems, government systems which are going to validate your information that you have provided. All the systems become part of a system of integration. Now when you talk about all these building blocks, what they are trying to achieve is just a platform which can cater to the various customer needs and requirements. And if you go a deeper dive into it, there are a lot of features and capabilities these platform requires, right, -- or each building block will require. For example, if you are talking about a system of interaction, you would need an API gateway. You would need even streaming capability. You would need capability to do an orchestration and transformation on top of it, right? Same goes for your system of record, system of innovation as well. Now if you have been working with Salesforce or we have a good understanding of Salesforce, you would already know that we have solutions that cater to each and every building block offer and to deliver that kind of a premium experience. But today, we're going to talk about very specific products. We're going to talk about a few Heroku and Data Cloud, right? But before we jump into those details, let's take a look of how various offerings of Salesforce will be curated together to deliver experience. Now you might have experienced layers, which could be in the form of a Tableau or a CRM Analytics. You could have mobile apps, custom or through a mobile publisher, we are doing it. You could have web apps created through experience cloud or a custom mobile app deployed on Heroku or you could have marketing or e-commerce in gen on top of, as an experience there, right? And then all these experience layers typically sits on your business logic. The business logic could be serverless function that you might have. It could be custom apps and EPS that you've built on Heroku. It could be sales force flows or a business process that we have implemented there using a local platform. And then for integration and action labels, we have MuleSoft as well. We have events-driven integration. We have custom processing written on or maybe data synchronization tools like Heroku Connect. We also have our own AI that is Einstein. And then, of course, Data Cloud is also goes there, which is a kind of a platform which generates all the insights across various systems and data that you have. And the data itself may be spread across multiple Salesforce offering. It would be in the sales of Service Cloud. It would be in the -- we might be using Lightning platform itself or you might be using commerce or marketing cloud as well, right? Now when we are trying to deliver these kind of apps, we often target a customer or partner or 1 app or 2 app or 3 apps, right? The way which kind of envisage a long-term strategy for our customer is that they should be able to deliver various persona-based apps through the platform, right? And when we talk about these persona-based app, we have introduced what we call as a Heroku mBaaS or mobile back-end as a service, right? Heroku platform being a Platform as a Service allows you to have all your various business functionalities exposed in your service. The overall approach is to build that, I would say, a repository of APIs that can deliver each of your business functionality in the form of a service. It could be simple as authentication or a log-in as a service. It could be our offerings or product catalog. It could be in the form of a booking or maybe, let's say, in the form of servicing or maybe per payment. Now these services will necessarily build as an experienced API layer on top of your various systems. So you will have maybe some Salesforce systems in place, like Sales Cloud, Service Cloud or it could be an ERP or maybe a customer build app that you have overtaken. But the whole purpose is to knit all these systems together and create a uniform layer where each and every business functionality is exposed as a service. Now when you have that, on top of it, what we can design is a multiple apps based on customers' persona. For example, you might have a customer app, especially targeting your end customers. You might have an influencer app for your influencers to say if you are, for example, a Century Plywood company that we work with. And then influencer app is important for delivering that experience to the architects, right? For a medical company that we work with, for example, [indiscernible] doctors could be the influencers, right? Similarly, there will be a distributor of Atara very important, especially in the retail industry, where you would have various distributor delivering your services, delivering your products to the customers. And then, of course, retailer or maybe an agenda. Some companies does have agents on payroll. Contractor region is on a payroll or agent on a payroll. For both cases, there could be a retailer agent app that can take care of it. But to enable, again, we would need that back end and that back end would be a mesh of APIs and that API is -- we propose to be in a form of individual services, which delivers that experienced API as a back-end as a service. Now how do we do it on a Heroku. So when we talk about a Heroku, it's just a simple platform, which act as a layer on top of Salesforce. In most of the scenarios, when the customers are using Salesforce products, they are using sales and Service Cloud most probably. And oftentimes, these services that you want to offer to your customers or partner, for example, you want to take an order, right, a retailer or a service agent, let's say, for example, service agent moves on road and takes an order or you want to build a customer app where customer can raise a request for a visit or they can raise a complaint. Now all this data is going to come and set into Salesforce, or it could come and set into your ERP systems, right? The way we envisage our Heroku is we envisage it as a layer on top of your back-end system, especially with the Heroku working very seamlessly with Salesforce platform and acting as a back-end as a service layer that can cater to various mobile and web apps that you can deliver on top of that, right? So let's understand what Heroku is all about, right? So you would have seen this customer just a lot of times, right? We again refer to it now and then when we deliver all our production placed in a wheel, right, so that we can cater to various customer-centric requirements, right? They could be sales and service and marketing cloud around it, right? Now all of this are achieved through a power of what we call the Salesforce platform. So when we call a Salesforce platform, and this is a Lightning platform we are referring to. Most of our core products -- in fact, all of our core products, maybe sales, service or even loyalty, for example, are created on top of this Lightning platform. This is a local platform, which allows you to deliver these great capabilities. It has inbuilt local capabilities like you have worked with flow, you might have worked with process builders. You might have worked previously with all this kind of an automation system that we have like approval processes, validations, object designers, report designers, dashboards, right? So it's a single integrated platform which delivers you all these capabilities in one package. But oftentimes, customer does create these capabilities by stitching various products and solutions together. For example, they might need a BI tool for reports and analytics. At the same time, they would have separate -- a workflow engine to carry out the workflows, and then they have a database to manage all their data, then they have another platform to design the presentation there or front end, right? In a Salesforce platform, you can deliver all that in one single platform. And it has a capability which can scale to any size of a customer. You must be aware that Salesforce is used by start-ups. It's used by mid-market segments. It's used by enterprise. It's used by Fortune 500 companies. So there is no, I would say, scale of transactions or a load that we have not had, right? What platform brings is, one, definitely, it brings you capability to faster go to market because of its local capabilities, readymade offering like sales and Service Cloud. So it increases the productivity of your IT team to deliver these services very rapidly. We have observed 28% increase in developer productivity when these business processes are automated and created in our platform. And then it also allows you to have continuous innovation. The reason for that -- one of the reasons for that is in a Salesforce platform, you get, what, 3 releases every year. And every release brings new features in terms of security, it brings new features in terms of AI, it brings new features in terms of ease of use, right? And that's where the next point is the complexity, we try to minimize it. We try to simplify each and every step, not only from the customer or end user point of view, but also from the administration point of view. Most of our products and services are configured through a wizard. So there's a getting started page, you go on there and say next, next, next and set up all this configuration. And with that comes a lot of responsibility, because when you are building all these business processes in, you will be managing a lot of data. And the data could be a company confidential data, data could be a sensitive data. It could be a customer's data, which would be PII data in nature, right? So security becomes an ingrained aspect of the platform. So whenever we try to evolve ourselves or think of a new product or services, security becomes a forefront of it, right? So we take care of security first and then we design our processes and we decide how it will be exposed to our customers, right? Now this platform, as you must be aware, is a SaaS platform. It's user-based platform. So every user would require a license. Typically, our Salesforce platform is either expose a per user license or a per log-in license, in which case you would have an Experience Cloud and you will be exposing these services to your partners or your customers and you have a log-in base prices. But both of the scenario requires users to be assigned licenses, right? Now this in itself is a good capability because it exposes a lot of features, but especially for the customers where there is a huge amount of traffic, where they're expecting their end customers to be like in tens of thousands of millions when they're expecting huge traffic and spike and flexibility in a load, then we have one more solution that we call as Heroku. So when we talk about Heroku is not a SaaS, is a platform as a service. It provides you capabilities to quickly deploy and run and manage your customer applications. And by doing so, it simplifies and takes away all the burden in terms of management, administration, infrastructure and things like that. At the same time, Heroku also provides you enterprise gate security. It provides you scalability on demand. That means if you have a very flexible load, let's say, one of our customers have a scenario where they receive a lot of payments in the starting of a month. They are NBFC companies again, and they receive a lot of payments in the starting of a month, 1st to 5th of every month. And when this happens, there needs to be a huge amount of flexibility in the system to cater to that incoming load. And that's where Heroku -- power of Heroku comes into picture, right? Now when we talk about Heroku, often people consider and compare it with the Experience Cloud. And it's a fair comparison because both products and services are created to focus nonemployees, right, for example, partners and customers -- end customers, right? There are various scenarios where one would be preferred over the other. We can go into a lot more discussion and debate about it. But typically, what we have observed that when a customer is looking for a very large user base, right, per user license model becomes kind of commercially viable for them, right? And also, when they have unpredictable note when they say that we might receive a huge amount of traffic someday because of our campaign or something, then in those kind of scenarios, Heroku becomes an ideal platform for that because it's a capacity-based model where you can provision the capacity that you need and you can scale up and down as you go, right? Now having said that, let's understand what is Heroku all about. So Heroku is PaaS platform, as I said, is a polygon platform, it understands multiple programming languages, right? So you build your apps in the language of your choice, you might be creating React Native apps, Angular apps, Node.js APIs or you might have Java or C# ,.NET programming language that you use. So you focus on developing apps, and Heroku takes away the burden of running and deploying those apps. And complete management, scaling, infrastructure, networking, load balancing is taken care by Heroku, right? Now the whole purpose or, I would say, vision for Heroku was that we give the power to developers, right? We give more and more capabilities to developers. So let's say you have an IT team. Now if you are working with any other cloud platforms, you might say that, I can definitely do this on AWS or Azure or Google Cloud platform. And that's true, yes. But typically, what it takes is it will take roughly around 2 to 3 weeks to start up a landing zone, right? Then you will require people who has understanding of infrastructure, right? Let's say you are working on AWS, you would need guys capable enough to manage AWS, create your VPCs, set up your EC2, set up a load balancing on top of it, expose the endpoint services, configure the firewall and BaFF, all these efforts goes into it, right? So the focus, which should ideally be on delivering the customer experience and creating those innovative user journeys, gets diverted and a lot of energy and effort goes into managing and running the platform itself, right? So if let's say you have a team of 20 people as an IT team, maybe 30%, 40% of that bandwidth goes just to manage and run the cloud platform, right? So when we say that we use Heroku instead, what we are bringing in is your worries of that 30%, 40% bandwidth goes up. That means all those activities, including management of infrastructure, including patching of your servers and management of your database, everything is taken care by Heroku. And what you worry about is your core, is the journey, is the experience, is the UI that you want to deliver. You focus on that. We focus on the business logic, right? So you write code and all the other activities right from the setup to deployment to scaling, everything which is related to infrastructure and management is taken care by Heroku platform, right? So you -- as a company, what we are trying to do is trying to spend more energy towards innovation rather than doing the routine and mundane jobs. And that's where the Heroku power lies. So if we go into a little more details, let's say, if you are setting up a -- you want to deploy our customer API, right? The customer basically provides you a loan origination process through our APIs and the data goes and sets on the sales force. Salesforce has a well-defined loan origination process that we have implemented. Now you want to enable your, let's say, partners and dealers to do that loan origination. And for that, you are designing those APIs, you can do it on AWS, Azure or any where you want, right? But whenever you do on any other platform, you will be probably using Kubernetes of any other microservices architecture to deploy and host those services, right? And Kubernetes does take care of a lot of worries. For example, if you're using a managed Kubernetes from AWS like AKS -- or EKS, I'm sorry. It takes care of a lot of worries in terms of your virtualization, networking and storage, right? But still, you have to manage your slaves. When I say slaves, I mean the slave nodes that you will add to MVC tool. You have to provision and size it. You have to manage your [indiscernible] files to understand how many ports I will be running for my container, right? And you will have the responsibility of managing the Kubernetes as well even though it's hopefully managed service from, let's say, AWS, you will still have to manage how many ports you want to run, how many kind of connectivity, how would you establish between the microservices, how they will talk to each, right? And if you're doing [ Beanstalk ] or if you're squarely putting on EC2 instance, you still have to take care of OS and security and orchestration and monitoring and all those aspects. When we -- you compare an experience with Heroku, typically all that worries goes away. We have a similar kind of a service that we call dynos, which provides you the capability to run containers, right? And all you have to do is click connect your core and click to deploy. Once you deploy, the scaling, the management, the running, load balancing, networking, everything is taken care by Heroku itself, right? So it takes away all those worries and lets you focus on core. Now what are the key benefits that you see? Of course, you are now focusing on [indiscernible] rather than developing your applications and then worrying about your management and upscaling. And then the most important part is that Heroku gives you a very seamless connectivity with the Salesforce. We have a component that we call as Heroku Correct, which synchronizes the data with the Salesforce. And what we're necessarily doing is basically doing CRUD operations on top of our database rather than making complex API calls of sales force and integrating with that area. It also allows you to have scalability on demand. So you could scale as big and as long as you want and not to worry about going in the mids or anything like that because it's a capacity-based model. You can plan to have 1,000 users at a time on a system or you might have 10,000 systems or 10,000 concurrent users at time on a system, right? It also provides you enterprise create high availability and security so that there is a, what you call it, redundancy inbuilt. So the high availability is in excess of 4 9's, 99.99%. It also takes care of your disaster recoveries, your VCP server. So all these things are taken care of. And that allows you to have a faster go-to-market because if you see in today's world, the digital processes or digital experience that are being innovated, you are running against time. Each business wants the process to be implemented yesterday. If you don't do it today, your competitor will go ahead of you, right? So it's very important to deliver those services quickly, securely and safely. And that's where the Heroku lies, right? So let's understand what does it offer. So it does offer a modern language framework. That means Heroku can understand your nodes, your reacts, it can understand your MVP or MVC or any kind of an architecture that you have. It provides you Kubernetes. It provides you containerizations. It provides you orchestration. So all those -- the capabilities that you would require in terms of event-driven or microservices architecture is taken care of. It provides you a comprehensive monitoring and scale. So you can monitor the KPIs that you want for example, throughput, you want to say a response time or I want to say a number of errors per second. All these KPIs and microservices you can monitor. It also gives you detailed monitoring logs as well and logging capabilities. So you have log streaming inbuilt, so you can do a very effective debugging and management through it. There is also managed data and other add-ons available. For example, you might want to use a Postgre database or a Redis cache or for example, if you want to use a Kafka. All these things are provided fully managed. Again, when I say fully managed, you get instance created in like, what, 5 minutes or something, and then you can start using it without bothering to worry about how to manage it, how to run it, how to configure it, right? And then, of course, we have a premium support. We also have practices and supports, which suggests you how to run this application, deploy those applications, what are the best architecture patterns that you should follow, in which scenarios, what capabilities will help you assist, right? So all these information are available to you here. Now when we talk about how do you perceive various components of Heroku. Basically, they can be easily grouped into 4 categories, right? First one is Heroku dynos. Any compute, any load, any processing that you want to do will run on dynos. Dynos comes in a various shape accounts. It could be a web dyno where you will apply any process that takes care of HTTP request, it could be API or it could a web app or it could have worker dynos, where you're doing some kind of a batch processing, maybe some background jobs you're running, you want some scheduled jobs, all those things will go in your worker dynos. You could also have one-off dynos. One-off dynos are short-lived dynos, which are used for maintenance job. For example, you want to create a slave database and take a backup of your data. You can run that on a one-off dynos or for example, you want to export a large amount of data and process it and ship it to some other location. That again is a good use case for a one-off dyno. Then you have Heroku add-ons. Heroku add-ons contains all these components we discussed earlier, like Postgre and Redis and Kafka and logging and [ New Relic ]. All these capabilities and components are available to fully managed through a dyno. And then we also have Heroku Connect. So Heroku Connect is that platform, which provides a synchronization framework with the Salesforce. And it's a very simple point and click tool where you configure which Salesforce objects you want to interact with. So let's say you want to have account contract leads and opportunity and you configure these other objects I want to configure, just click, click, click, save, and it sets up a bidirectional synchronization with your Salesforce. So what it necessarily does, it creates a database tables where all your account lead opportunity data will get replicated. And on the Heroku side, you write APIs and business logic that will interact with those tables. Now you are just doing read, write operations on these database table. And when you read the data, you will get the latest data available from Salesforce, but you don't have to call any APIs just to read the data from table. When you want to update the data or insert the data, you are just making insert and update on the tables and Heroku Connect will push the data into Salesforce. So the synchronization of responsibility between the tables and the Salesforce objects is taken care by Heroku Connect. And all you have to worry about, all you have to focus on is interacting with the database, which, I mean, is a very simple job as compared to calling those complex APIs and then stitching the journey by creating the API calls from Salesforce, right? And last but not the least, there is a component that we call Heroku Private Space. We also have another, I would say, version or more secure version of our Shield. So Private Space is basically a private network or a VPC that you configure. Typically, for an app, you might want to have 2 private spaces. For example, I want all my APIs and database to be in one private space and my web apps or my customer-facing apps to be on another private space. So private space is nothing but just a local segregation of how your assets or workloads would be segregated. Oftentimes, our private space is also used for managing all your ease-to-use communication. For example, you might want to have a service fabric where your multiple microservices can talk to each other. Having them in a private space enable that ease-to-use communication between APIs and microservices as well, right? So it takes care of highly secure environment, and it makes sure that you can protect and block the traffic coming in and out of that private space, right? Now if you look at the overall architecture from the from the, let's say, 10,000 feet layer, you would see that you have -- at the top, you would have corrected experience, created as it may be in the form of apps or mobile apps or APIs. Now they could be very well running on top of Heroku or you could be working with the Salesforce. Now Heroku allows you to extend those capabilities as well. There are a lot of popular use cases where Heroku is used to take care of, for example, processing of a Salesforce data. People use it for archival. People use it for running large processing and batches, which they are not able to or maybe facing challenges to run on Salesforce or at times you want to have some maybe open source libraries that you want to use for the business processing and the logic that you want to write. And that allows -- that again, Heroku enables you to do that, so you can use Python libraries and white programming that can still work on data of Salesforce, right? And then, of course, all the administration, all the security, all the DevOps capabilities are taken care of. So let's take a quick look on how it will work. So typically, the way it works is that you have -- one second, yes. So you -- a developer writes a code, simple, so you are writing a code, you might be using Node or PHP or Python or whatever the preference programming language o scripting language you have. You write that code. And then you connect that code probably using a GitHub or any kind of maybe a CI platform that you use. So whatever repository that you have where your code will store, you get correct that repository to Heroku. What Heroku does, it is smart enough to understand first what kind of a script or programming language or a Javascript you're using. And it automatically picks that code. It understand that what is the right compiler to package this code. So it compiles that code. It builds the code and build a package image, which we call as slug. And that slug is something that translates, is equivalent to a container image. And that slug runs into the dyno. And dyno, you can say, is built-in container, right? So it automatically does all this part. All you have to do is connect your repository to Heroku. It understand the codes. It understand the programming language. It automatically packages the code. It creates a slug or a container image and it start deploying that into dynos and it starts running it. And once you do that, you have -- it has access to the data service that you will add to it. For example, we have deployed the APIs, which want database. You want maybe have a Redis cache, which you use for cache and session management or you might be interacting with Kafka or even bus, all these data services you can have with it. And then there are more than 180 plus third-party add-ons that you can have. There are add-ons for designing reports, there are add-ons for managing identity platforms. There is add-ons for processing files, storing large amount of files, all these add-ons are available to you, fully managed, right? So user access your code -- I'm sorry, user access your app directly. There is probably, I would say, 15 minutes, it will take from writing a code. After writing a code till a production are app, it takes 10 to 15 minutes and you're live with your customers, that fast how it is, right? And then runtime, orchestration, all these capabilities are understood and managed by Heroku itself. You don't have to configure any ML file or write, how many codes I'm going to write. It's a very simple write drop configuration that you can control and manage. And you can also choose to go with the Autoscale. We'll see how it works, right? And then we also quickly want to touch upon another offering that we have that's called Salesforce Data Cloud. Now how Salesforce Data Cloud fits in all the scheme of things. Now oftentimes, the experience you want to deliver or a data that you want to show in a front end is not limited to Salesforce itself. You might have data which is spread across multiple external systems. For example, we were working with one customer which works on home loan, which is a home loan provider. They have fixed deposits. They have NRI loans. So when they are delivering an app to a customer, they want to expose all these services to it. And the data is spread across their loan management system, Salesforce CRM, fixed deposit system and other systems, right? Now in this kind of a scenario, they can often -- of course, a lot of time, right, point-to-point integration with these various systems. They will have integration with the loan management system or an FD system. But a better way to do it, a simpler way to do it using a Salesforce Data Cloud. So Data Cloud is basically our data platform, which allows you to necessarily connect to your various data sources or various system. They could be cloud-hosted systems, for example, Salesforce or it could be the data which is stored on S3 bucket. It could be on-premise database very well. It could be very well at that. Or it could be hits or interactions that you are getting from your APIs. It could be your IoT system. So you can bring and connect all the data sources in, harmonize that data into your target data model or canonical data model. It can also help you unify your customer to create a single customer profile. When I say a single customer profile, let's say, you have a customer which is -- which has bought a product, right? And that's why you have a -- his profile in a CRM. Now the customer comes to your service. Let's say, you are using ServiceNow there, and it's created a case. In that time, he has given you a different e-mail address or phone number or name has been captured with a different -- slightly different spelling. So now you have 2 profiles of a customer. One in Salesforce CRM, another in ServiceNow, right? Probably also have a record in your marketing systems and let's say, Adobe. In Adobe marketing system, you have another record for a customer where you are sending in promotions and e-mails, right? What Salesforce Data Cloud enables you to do is you can connect through all these 3 systems together and then kind of unify this record and identify one single copy of a customer record. So you are necessarily merging these 3 customer record from Salesforce CRM, ServiceNow and Adobe Marketing as one customer. And in a nutshell, you are -- and you're reaching that profile because there are different elements of data which are setting in different systems. And by creating all this unification rules, you are basically creating a golden copy of a record of a customer. So one record of a customer, which now has all the information, which might be spread across these 3 systems, right? Once you do that, you can do a better segmentation, better targeting for that customer. And you can understand your customers better. And when you are running your ML models, for example, product recommendation for that customer, it will be more, I would say, effective, more -- the efficacy will go up, right? So Data Cloud allows you to, again, connect to these data sources, bring them into a very well-defined structured data models and then run an identification rules to unify your customer data, if you want. And then on top of it, you can do analysis, you can do all your AI, BI. And then also apart from AI, BI, you can act on our data. So when I say act on our data, for example, let's say, a customer is going on a website and browsing through your page through an act when this data, which can come both as a batch and a streaming. So interaction on our website will probably come as a streaming, right? So when the information comes that customer is watching a video on your website for 10 minutes of a product, you can add the customer to a segment or probably create automatically a leader in Salesforce, which says that this customer can be targeted for this particular product, right? So all these information, for example, let's say, a customer has made a last payment of his loan, we can start giving him -- for example, he has completed a home loan. You can start giving him in terms of home improvement loans, right? So all these actions you can take in the Salesforce through the power of flow and the Data Cloud. But it also provides a capability, we call Webhooks. And using Webhooks you can also expose these signals to an external system, right? So again, what we are trying to do is trying to understand data by connecting and ingesting data from various data sources, making a sense out of it, creating insights, running your AI models, all you can do in one platform called Data Cloud and then take an action on top of it, right? So that's where the power of Data Cloud is. And when we create those kind of an experience, the customers that we saw earlier, now become a lot more AI enabled. Because AI can only be effective if you have a good amount of data. And data is often spread across multiple systems. So when you apply Data Cloud, all your data gets consolidated and it makes much more sense. So you have maybe better pipelines, better experience to your service agents. You are delivering and closing cases faster. You're improving your ROI. All these capabilities, you can deliver through Data Cloud, right? Now next, what we're going to do is we're going to go through a quick demo to see how Heroku and Data Cloud and maybe other custom apps and platforms can come together and deliver the customer experience, right? So let's take a look. So what we have is a hypothetical company called Radiant Golds and Diamonds. Now Radiant Golds and Diamond, as the name suggests, is a jewelry company, and they have created a client training app. So this app is for their sales agents, which often works in various showroom. And since Radiant is a premium jewelry brand, they are working with high net worth individuals. So the experience they are going to deliver through their various channels. And most importantly, jewelries often what, especially the high price is often what in showroom. So they have to deliver that premium physical channels as well, right? So the client training app they have created is meant to deliver their experience through your sales agent to the customers, right? So this app is for sales agents again. And we're going to take down a person of sales agent called, Ankur -- Arjun, I'm sorry, to deliver that experience and how his life looks like, a typical day looks like. But before we do that, as a developer, how Heroku can empower such an app, right? So as a developer, what you typically do is -- this is what Heroku console looks like. So you are ready with your application. You have created this very nice tablet app on React Native and your APIs are running on Node.js and your data is on Postgre, which is synchronizing with Salesforce. So all this is ready with you. Your code is ready, you have done developed. And now you are deploying the staging of production. So what you have to do is the first step is connect your repository. You can use Heroku's in-build git repository or you can have Github or any other kind of repository you might have. And so you will connect that repository, you'll specify which particular project or a repository you want to connect to. And once you do that, you have an option to either choose to do automatic deploy. Oftentimes, you will do automatic deploys and staging. So every time you commit your changes in a main branch, the code get automatically deploys and you don't have to worry about building and packaging, right? So you can very easily enable automatic deploys and set up your testing suite as well, right? So it will do a sanity check every time there's a deployment done. So the way it works is just enable the automatic deploy. What you're going to do is checking your code, your latest version gets check-in into main branch or branch of your choice, for example. And when you cut check-in the code, automatically deploy will -- it will detect automatically that code has been checked. And the deployment will happen, right? For the first time, let's say, you might choose to do a manual deploy. So when you click on a manual deploy, the Heroku will go and tap into your repository. And then it will read that code, and it will understand what is this code all about. This code is written on whether Node or Java or C#, and it automatically identifies how to build and package this code. You don't have to specify anything. Just connect to your code, it will automatically detect and understand, and that it will automatically select a build pack, which is nothing but a set of steps and scripts to package this code. So this build starts happening automatically, and then it will package the code and deploy it. So once it is deployed, your application up and running. Now typically, once you deploy it, you would try to see how much compute I would need, how much -- what should be -- let's say, I'm expecting 10,000 users per day, right? So what should be the size of my compute on which this app could run. So it's pretty simple. You don't have to configure any [ gamble ] file to define a number of nodes. You can just click on the number of dynos and you can upscale by choosing more powerful dyno, for example, performance and in this case. And you can also choose to increase the number of units by default for production, we suggest you run at least 2 dynos to give you that kind of high availability. But you can choose more. So if you anticipate more load, you can manually just click and drag this bar or you can put that number of dynos you want to run over there. So let's say, we choose 4 and go ahead with that. You also have an option of auto scaling. So when you choose auto scaling, you are basically configuring a setting which tells Heroku survey my dynos numbers between 1, 2, let's say, 10 based on response time of 1,000 millisecond. So what does it necessarily mean that if 95% of my response of my API or web requests, whatever it is, if 95 percentile of that response is within a threshold of 1,000 milliseconds, then I'm good. I don't need to auto scale, right? But if the load increase -- when the load increases, the performance and the response time goes down, right? So sorry, performance goes down and the response times increases. So when the load increases and response goes beyond 1,000 milliseconds for more than 95 percentile of requests, it start adding more dynos. So it adds more dynos, analyze the data for more time. And if still, the response time is not within 1,000 millisecond threshold, it adds another dyno, right? So that's how it works. And it will keep adding to the -- till the maximum rate that you have specified; in this, it's 10. And when the load goes down and the response time increases and stays within the limit of 1,000 milliseconds, it start decreasing the dynos. So it will down scale and bring it to the minimum threshold that we have defined; in this case, 1. So it will fluctuate 1 to 10 based on response time of your APIs or web request, right? Once you configure it, you don't have to look back. So it saves you in terms of -- if you, let's say, you can choose to have 4 dynos running always. But that's not a very cost-effective way to do things, right? Autoscaling will save you a lot of dollars. When you don't have load, it will come down to 1 or 2 dynos based on your configuration. And when you have load, it will upscale to require numbers at the maximum threshold, and you can also choose to be notified when the threshold reaches, right? So by that, you can do an autoscaling and is as simple as that. You don't have to write any scripts. You don't have to write or update any external files for that, right? Now let's step into this person, which is, I think, Arjun. I'm logged in as Arjun, which is an employee in one of the stores of Radiant. Now when he logs in, he sees this window. Now this window gives all the information that Arjun needs on its steps. It tells him where he stands in terms of his targets versus achievements. It tells him where he stands in terms of the leaderboards, are there any campaigns running. It also tells him all about his task, connectivities, what he needs to do today. So he can see that there's a team meeting schedule. He can see that there is a store visit planned for Shreya, right? So Shreya is a long-standing customer of Radiant and Arjun anticipate her visit today. So let's see how Arjun will meet and target Shreya. So typically, if a smart sales guy would be there, he would understand what she is all about. So he opens Shreya's profile. In Shreya's profile, he sees a lot of details, right? Now all these details cannot be from one system. It's not. It's practically not possible to maintain all the data in CRM, right? And this -- the information which you are seeing right now is coming across multiple sources. At the top that you see, the section is more of a profile of Shreya, which is coming from CRM. Now this data, which includes her date of birth and with e-mail address, could be spread across your multiple system. For example, it could be in your lead management system, it could be in your marketing system, it could be in your service -- customers server systems. With the power of Data Cloud, the unified golden record of customer that you have created, you get all this information in one place, right? So you get this data -- complete customer profile coming on top of it, right? So all this information is available to you. Then you see a lot more insights like engagement scores. The engagement score is a measure of interaction. Shreya has been doing through your various social and marketing channels, right? Now all this information is coming in Data Cloud, and you have written a logic to calculate and engage in storefront. She is also -- has been a long-standing customer. So you have calculated various insights that she has a lifetime value, what is the last lifetime purchase value, what is the size of purchase that he has made -- she has made on life, how many loyalty points she has connected, what is her most recent store visit plan, right, along with what her choice of jewelry is, right? You also see the loyalty information over there, which includes that she is a part of golden tier, and she's eligible for a special birthday offer, which Einstein has recommended for her. Then you also see what various activities she has been doing probably on your e-commerce portal, on your website, on your marketing. So let's say, she opened an e-mail. You can see that. She has done a product browsing and you can see that. And you can also see what are the recent products that she has watched -- she has seen browsing on your website. Now when she has done that, it allows you to make a very uniform choice. Now Arjun has the information about her taste, her profile, her preference. He has also had a recommendation from the Einstein. He has the information about her recent activities. And based on that, the most recommended products are listed over there. Einstein has also calculated, if you see over here in the recently products, a propensity to buy, right? So this propensity by scores, does the calculation based on machine learning and say that this -- there's a likelihood that a customer will purchase this product. So now when Shreya visits, Arjun is already equipped with all this information. So what he does is he picks that these are the 2 things or 3 things I want to kind of select for Shreya and when she comes, I want to show it to her, right, or recommend it to her. And when he does that, Shreya could get very well impressed through all these processes. And then she tries it on. She tries the rings, chooses a size and decide to make a purchase. And that's how Arjun will be able to deliver that premium experience to Shreya and Shreya will be a happy customer, right? And Arjun can do much more than that. The experience doesn't stop there. It doesn't stop with the physical channel. Then Arjun goes and logs into her campaigns -- his campaigns, and he sees that all these campaigns are available. So he picks one and tries to share those details and say that I'm going to share these scheme details with my customers. When he clicks on the share button, he sees all the customers that he has. Not only that, but he also sees the preferences these customers have. Some customers prefer to have e-mail, some prefers SMS, some perfers calls, right? And when he sees this information as for the preference of a customer, he can now target each customer based on their communication preference. And he does exactly that. He picks the customer who have preferred to be contacted on e-mail or SMS and sends outs promotion discounts to those customers, right? So this is how you can deliver that premium experience not only through digital channels, not only through marketing channels but also the physical channels. And by having the power of Data Cloud to consolidate this data across various sources and generate those insights can make you -- can enable your employees and customers to make a lot of informed decisions. And with the Heroku, you can provide them a very premium, very brand-centric, very customizable digital experience to them, which can elevate their overall journey or experience or, let's say, perception of a bank, okay? And the power of Heroku also allows you to scale it up and down, cater to the demands in terms of variable flexibility, which you cannot predict, and allows you to cater to various large audience, very huge size of audience without worrying about the per user licensing model. [indiscernible] licensing allows you to do that. And that's all we have for today. I'll quickly go through the Q&A that you might have.

Nikhil Dhruv, is there any specific question that we should target? I think most of that has been addressed. Can we create a developer version of Heroku? I think Nikhil has already answered the question. I'm just quickly checking. [indiscernible] ROI calculation. Yes. So we will do that. I think there is a reference material that you will get from our team. In those, you will see certain links coming. You go through those links, it gives you very popular use cases and understanding over there. And you can use those architecture principle to calculate ROI as well, right? Okay. Okay. I would like to address this question as well. Nikhil has already answered that. ROI of Heroku. Now -- yes, again, as I said, most of the cases Heroku will be compared with the Experience Cloud because through power of Heroku, you are trying to target nonemployees or users whom you don't want to kind of give up per user license model, right? The power of Heroku effectively comes from 2 places. One, it comes from providing seamless connectivity to Salesforce data. And number two, the automation that brings, right? So when you are calculating any ROI, you can completely get rid of your infrastructure team when you are using Heroku. The developers could easily work with Heroku without much of a training or experience and run the production grid applications. So that's where the power of Heroku basically lies, right? And along with that, if you're comparing with Experience Cloud, Heroku will make more sense as the number of users increase. So let's say, if you are targeting an app for a partner, then your partner base is that's 1,000, Heroku might not make too much sense in terms of commercial values. But let's say if you are targeting a customer which has partners in excess of 10,000, then you will start seeing the value of Heroku. As the number of users goes up, the per user cost of delivering that experience will go down, will continue to go down. So that's where the Heroku power lies. Will you be sharing this PPT with us? Yes, I think we can. Any other questions? How can I get updated with the Heroku market base? Is there are any news letter on Trailhead? Yes. So you can go to elements.heroku.com. We're going to send those links to you. You can go through that. Trailhead, yes, trail is a very good source for it. Unfortunately, right now, we do not have too much of material of Heroku in Trailhead. We are working on creating more assets, more modules and bargain on Trailhead. So you will start seeing lot more activities on Trailhead. But right now, the best space will be devcenter.heroku.com. We'll share those links with you or elements.heroku.com, where all these elements and plug-ins participate over there. Can we get a demo exercise for the trial and understand the strength of Heroku? Yes, Heroku does have a very easy way to onboard. It's now available with the credit card. It has a lot of free features available. There is also a dyno type called Eco dyno, which automatically shut downs when you are not using it. This is a free tier for Heroku. There's a free tier for database, free tier for Redis, yes, you can definitely do that. So all you have to do is just create an account and start using it. There are a lot of free offerings which are available that you can start using. How do you ensure the security and performance of application in Heroku? Yes, good question. So when we talk about security and performance, there are 2 main things that goes into it. First is the redundancy. Now if you are choosing a free tier, you cannot really expect it to have kind of security and availability. So when you're choosing a particular asset, for example, you're choosing a database. What -- necessarily what you need to do is choose the right plan. You will see that there are multiple plans listed over here, which gives you various levels of security. For example, you want to have high availability, which is in excess of 9 4's, and you also want database encryption. So you get a premium tier and you get it. If you want a database, which should be HIPAA compliance, there is a tier associated with. For example, you will go with Shield or you will go with Private Space. So it depends on the tier of plan that you choose and all the security and performance criteria are digged into it. And how do we ensure that? We ensure it, for example, for a database, pretty simple, we deploy the high availability architecture. There is a team of some engineers who monitors them, who make sure that your applications and databases are up and running. Your infrastructure has no issues. So there is a inbuilt redundancies and our team, production support team, which takes care of uptime and availability of infrastructure, managing, patching, all these things are taken care of -- with switching your workload at the appropriate time. So you might see a notification that the database will go up, will go through an upgrade. It will give you at what time this will happen. And the way they do it is they first switch -- they first set up a new database with all your data synchronization and everything, and they switch your load and then upgrade your original one. So you have a switch and update kind of a model. So it minimizes the downtime to be very least possible that we can and have the performance measures and upgrades and security ensured short, right? There are -- for every asset, there is a different strategy, and that's how we ensure the security performance available, right? I think we are out of time now. If there are any outstanding questions, please do post and write to us, we'll come back to you. The deck and a lot of links will be shared and that will give you a lot of information about Heroku and Salesforce and all these updates. I hope you found this presentation useful, and we do hope to continue and give you more and more webinars like that. So please let us know your feedback and where we can improve further. Okay. Let me take this one last question. Can you explain the dyno and Heroku and how they're used to an application? Okay. So dyno is much like a container. So let's say, if you compare it with Kubernetes, in a Kubernetes, you have a master and you have slave loads. These are nothing but servers, which will take care of load. Master takes care of your management of time, scheduling and all that, commands and all that, and slave takes care of your workload, right? Similarly, in a dyno environment, you will have a Dyno Manager. Dyno Manager is nothing but like a master in Kubernetes. And then there are slaves where in which the dyno's containers are created. These are Linux containers only. Nothing special about it. So much like any containers you might have worked on, right? Like ACI, Azure container ,[indiscernible] you might have worked on or elastic cumulative service or AWS or Google Kubernetes service, right? All these are containerized environment. Kubernetes is one way of doing it, dynos are another way, but it is just a Linux container. And how they're used to run application? As I mentioned, you could have a web dyno, where you can deploy your web APIs or web apps, or you can have worker dynos, you can have one-off dynos. So various dynos can do a lot of jobs. They can run batch processes. They can run background jobs. They can run -- they can accept the HTTP request and provide a response, right? So that's how your dynos are used across our divisions. All right. Having said that, I'll wrap it now. Thank you so much for your time being. I hope you enjoy the session. Please do share your feedback. Thank you so much.