MuleSoft powered AI and Connecting Data Cloud with Mule

Editor's Notes

• #5: Key message: We’re in an AI revolution, and Salesforce is going to help customers navigate each new wave of AI. Now, we're in an AI revolution. But at Salesforce, we have a long history of helping customers transform with AI... We started in this first wave, predictive AI, with Salesforce Einstein. We knew all the way back in 2014 that AI was going to change customer relationships We established an in-house research team that helped pioneer AI for CRM with the launch of Einstein. Einstein has driven amazing results for our customers to sell smarter, serve smarter, and engage smarter. Earlier this year, we entered wave two, generative AI, with the release of our GPT products. You might not know that Salesforce researchers published some of the first papers on prompt engineering and generative AI. We’re already seeing incredible productivity and a range of new capabilities like deal insights, account summaries, briefings... And now, we’re rapidly entering a new wave of AI Soon, autonomous agents will be able to communicate with each other and do tasks for you. We’re already launching agent-driven experiences within Salesforce, which we’ll talk about more in a bit. As we move through these technology waves, CRM will become even more important to our customers than ever before And Salesforce is going to help you seize the extraordinary potential of AI.
• #6: Talk Track Organizations today are managing: A complex web of siloed systems and data. The average enterprise has data in 991 systems with only 28% integrated. Without connected data, our AI models don’t have the business context they need. When systems are integrated, it’s using a variety of tools, which leads to a complex, heterogeneous, expensive integration environment. Integration has been such a major roadblock that 95% of IT leaders report this challenge as a leading barrier to AI adoption. With sprawling data and APIs, managing and securing data at scale is even more difficult. This is why one quarter of all APIs go ungoverned. All of this makes it nearly impossible to orchestrate and protect your data in order to deliver on the promise of Generative AI. And with IT backlogs growing almost 40%, it’s no wonder that we don’t feel ready to adopt generative AI.
• #7: Talk Track: This is because AI needs to be grounded on data to generate valuable insights and personalized output, and it needs pathways to action these insights and outputs in relevant systems. However, most of the data and systems necessary to accomplish this are disconnected. 62% of data systems are not harmonized to leverage AI (Connectivity Benchmark Report, 2024). Integrating these systems takes a tremendous amount of time from your IT teams. Without a strategic approach, integrations can be expensive to maintain, they're brittle, and maybe most difficult of all they must follow the growing security, privacy, and compliance regulations emerging in the world.
• #8: Talk track: As we discussed, you have data spread across hundreds of systems: what if we could convert all of that data into composable building blocks, that allow us to innovate and move faster? How do you make your data work for you? First you need to unlock your data, from anywhere, and make it discoverable. This includes on-premises, hybrid, cloud data, as well as data in any format, including structured and unstructured data. Then you need to secure that data, with API policies, specific to your industry and company practices across every API in your digital estate. Then, you need to automate and orchestrate that data across any system, AI model, bot, or process to create intelligent experiences for your employees and customers. This is what we call the FULL lifecycle of a building block. Now, point solutions exist across integration, API management, automation, and AI to achieve some of these capabilities, but we’re seeing the market is converging to enable organizations to manage that full lifecycle more seamlessly. THIS is the key to achieving the agility we we need to adapt and compete in today’s market. And it all starts with integration, let’s take a closer look at what we mean.
• #9: Talk Track MuleSoft opens the universe of connectivity. Not only do we have hundreds of connectors for SaaS applications, ERPs, and legacy systems, but we also have one unified platform where customers can build, deploy, and manage all of their integrations.
• #10: Talk Track Together with Data Cloud, MuleSoft can: Enrich the data sources natively available with Data Cloud by connecting data on ANY system across any third-party, legacy, or hard-to-reach system like FIS or Epic, SAP, or mainframe systems. Action specific segments from Data Cloud and insights from Einstein by creating event-driven integrations and automations that communicate and interact with any downstream system in real time, for example, you can update records across your business and trigger multiple systems and workflows to fulfill customer orders. Enable visibility over and monitor all of your integration points from under a single pane of glass. With one solution to connect all outside data to the Einstein 1 Platform and action insights real time, you can provide the best possible customer experience as fast and efficiently as possible. On average, MuleSoft customers are able to go to market 78% faster and reduce maintenance costs by 74% by leveraging reusable APIs and integrations.
• #14: Talk Track: Let's walk through it together. We first connect to all of your data sources, whether that is your Customer 360 data, third party data from Google Cloud Storage, and Google BigQuery, web and mobile data, legacy systems and so much more. And we do this with an extensive library of out-of-the-box connectors. Once we ingest and bring all of this data into Data Cloud, we prepare and transform it into a new view. But the challenge is you can still have data in a variety of formats. For example, your customers are called contacts in sales and service cloud, while they are referred to as subscribers in marketing cloud. With clicks not code, we harmonize this data, no matter the format, into a common data model. And together these data models from a single customer graph or customer profile. This is the true 360 degree view of every customer. Now, this is my favorite part of how Data Cloud works. Because you can take this unified, real-time data and activate it across the entire Customer 360 and our third party partners. Which means the use cases across your organization are truly endless. You can power smarter AI with Einstein GPT. You can automate any manual process with confidence using Salesforce Flow. You can explore and visualize this data in Tableau to unlock real-time insights. Now, these are just a few examples. Transition: Put simply, Data Cloud helps you connect and harmonize all of your customer data. Then activate that unified data across the Customer 360 and trusted third party partners to personalize every customer interaction at scale.
• #15: Now let’s look at how Data Cloud and MuleSoft compliment each other in two main areas: ingestion and activation. Data Cloud has a number of native capabilities that can be extended and enhanced using MuleSoft. Ingestion: Data Cloud includes built in capabilities to ingest and federate data from: Any Salesforce org A growing list of widely used SaaS applications and hyperscalers (e.g. PaaS platforms like Azure, GCP, AWS) - we have announced 200+ native Data Cloud connectors Unstructured data from text, html, and PDF files Data lakes and warehouses with zero copy MuleSoft expands the types of systems and sources we can ingest into Data Cloud: Data from on-premise systems and accessed through firewalls - MuleSoft can run locally to stream this data to Data Cloud; MuleSoft can support any network architecture Data from transactional systems and ERPs that could benefit from queuing, error handling, and delivery controls Unstructured data from document repositories like Google Drive, Sharepoint, and Confluence and images like scanned PDF files, png, and jpg using OCR Activation: Data Cloud includes a number of egress capabilities to action on data and insights: Activate data across the C360 using Flow Builder and enrich CRM objects with reverse ETL Ability to share data from Data Cloud back out to data lakes/warehouses with zero copy Send targeted ads on platforms like Google and Facebook Visualize data in Tableau for future analysis and insights MuleSoft builds upon these activations with the ability to: Respond to data events outside of Salesforce, in external systems using APIs and bots Respond to a data event with automated workflows in Salesforce using Flow Orchestration (remember - Flow is part of the MuleSoft portfolio and your Mulely can help you position and sell Orchestration)
• #16: Key Message Ingest data at scale using native connector to over 160 applications Optionally use MuleSoft for applications not directly supported by Data Cloud or that need complex error handling, orchestration or alerting Talk Track Data Cloud comes with support to connect to, and extract data from over 160 systems. Some of these include Amazon S3, Google BigQuery, Postgres and more. Data Streams allow administrators to setup inbound connections and select data fields to import into Data Lake Objects. These inbound connections can be either scheduled (batch polling) or streamed (near real-time). For applications not directly supported by Data Cloud…or for more complex pre and post processing… customers can make use of MuleSoft’s Anypoint Platform. This comes with pre built connectors for Data Cloud and over 200 enterprise applications. Mule applications provide native support for complex routing, orchestration and monitoring. Transition With multiple sources of data the question becomes how does Data Cloud provide a common model to drive business insights. This is where mapping disparate data source structures to a canonical model is important.