Hyperautomation & AI/ML: Keys to Digital Transformation Success

Editor's Notes

• #4: Audience poll questions: What level of automation maturity has your company achieved?   Very high-level of automation (all processes are automated)   High-level of automation (most processes are automated. Very few manual processes)   Medium level of automation (some processes are automated, and some are manual processes)   Low level of automation (very few processes are automated, most are manual processes)   No automation maturity yet, but we are planning to implement.
• #5: What are we going to talk about today…
• #6: Why is automation such a hot topic in the IT world these days? The pandemic showed many of the shortcomings of our decades of digital transformation efforts. We have spent years and countless dollars digitizing our data and processes, but when the pandemic hot, many of us were caught flatfooted by it’s limitations. As a result, we are now seeing that for many of us, the next step on our digital transformation journey is automating our critical business processes. Many of these are centered on SAP ERP systems and involve creating and managing master data. Now, let’s look at the needs and wants that companies have coming out of the pandemic and how they can be addressed through automation.
• #7: Gartner – And growing at 12% CAGR KPMG KPMG ASUG Gartner
• #8: Talk track: Coming out of the pandemic, many of the critical needs our companies have are the same as they were before the pandemic, but now there is a more urgent need to address them before we are faced with another sudden, existential disruption. Talking with our customers we believe the three core needs that companies have, which can be addressed by automation are: 1.) The need to be more agile, with the ability to pivot and adapt when our businesses or markets are disrupted. 2.) We continue to need to go faster to meet competitive threats and keep up with the ever-accelerating changes we see in the marketplace 3.) We need to do these first two while improving the quality and integrity of our most important data – our master data. Since data has become the fuel we use to make critical business decisions, we need to make sure that it is of the highest quality and that we can rely on it.
• #9: This quote does a great job of summing up the importance of data in executing digital transformation. It sets up our discussion for the interdependence of process and data and sets Precisely apart from many of the other process automation providers.
• #11: Why is automation such a hot topic in the IT world these days? The pandemic showed many of the shortcomings of our decades of digital transformation efforts. We have spent years and countless dollars digitizing our data and processes, but when the pandemic hot, many of us were caught flatfooted by it’s limitations. As a result, we are now seeing that for many of us, the next step on our digital transformation journey is automating our critical business processes. Many of these are centered on SAP ERP systems and involve creating and managing master data. Now, let’s look at the needs and wants that companies have coming out of the pandemic and how they can be addressed through automation.
• #12: Gartner coined this term to represent a broader set of technologies, solutions and capabilities that were increasingly converging into a newer market than what used to be called the business process automation market.  Hyperautomation is a business-driven, disciplined approach that organizations use to rapidly identify, vet and automate as many business and IT processes as possible.   Note there is no single process automation technology that provides every capability required to deliver hyperautomation across an organization.  For instance process mining isn't the same as process automation yet both are a part of the hyper automation market. In many cases, selecting technologies is not about simply comparing one tool to another tool. Rather, customers are increasingly focused on determining the best combination of complementary technologies. Hyperautomation involves the orchestrated usage of multiple technologies.  This means customers have a toolbox for automation and we need to fight for the right to be not just one of the tools in their automation toolbox but the selected tool for a given business problem.
• #13: Remember what Gartner said about Hyperautomation: It involves the orchestrated use of multiple technologies, tools or platforms This is one of our customers that are currently using Precisely Automate as well as a generic RPA solution. In this case, it’s Blue Prism but there are other examples where customers have chosen at least 2 automation tools— Precisely Automate for SAP processes and an RPA bot provider for non-SAP processes. This was quoted during an RPA Automation Breakfast Hosted by PwC
• #14: Why is automation such a hot topic in the IT world these days? The pandemic showed many of the shortcomings of our decades of digital transformation efforts. We have spent years and countless dollars digitizing our data and processes, but when the pandemic hot, many of us were caught flatfooted by it’s limitations. As a result, we are now seeing that for many of us, the next step on our digital transformation journey is automating our critical business processes. Many of these are centered on SAP ERP systems and involve creating and managing master data. Now, let’s look at the needs and wants that companies have coming out of the pandemic and how they can be addressed through automation.
• #17: And the example we want to show you today is the creation of Material Master data in SAP. For Organizations typically in the manufacturing industries that have a high amount of new product introductions and changes in any given year, that means a high volume of new material requests and changes to existing material master data in SAP. Now these processes are highly manual and complex processes with lots of data entry error points. We see here an example webform to create new finished goods. It is not uncommon for organizations to have 2-3 hundred fields or more on such a material request form. This data typically has to be collected and entered by different teams and can come from various sources. It therefore usually requires multiple collection steps, validation steps, review and approval steps in the process.
• #18: Because of the volume and complexity of data, companies typically set out to define 100s if not 1000s of data validation rules that they implement. And that works great and our solution provides the flexibility to support any amount and any complexity of rules with ease. However, the process to define and agree on business rules in the first place is often a lengthy and cumbersome process.
• #24: Our Automate platform delivers a complete set of automation capabilities from data automation to complex process automation. Data automation enables you to automate mass data creation and management activities such as pricing or materials changes using Excel. On the process automation side Excel workflow enables you to manage the data automation files through a complex workflow. A prime example of Excel workflow is for journal entries in finance. There is a good chance that you are already managing your journal entries in Excel. Excel workflow allows you to manage these Excel files through a contingent review and approval process that uses your business rules to determine the routing and approval process. And once the final approval is received, the data is automatically posted to SAP. No manual data entry at the end of the process.
• #25: For example, if you are creating a new G/L account, the person who initiates the process enters basic account information in the form. Based on this information, and rules, that are defined and setup by the customer, the Precisely Automate solution determines who the appropriate contributors are, who needs to approve the final form and then can auto-post the approved data to SAP, thus creating the new account. No more manual routing processed, error-prone data entry or ungoverned files being shared via email.
• #26: One of the core, defining characteristics of Hyperautomation is the broad set of capabilities to look for when adopting a hyperautomation strategy. Let’s look at some of the featured capabilities that the Precisely Automate platform has which help it form the center of yout Hyperautomate strategy: - Workflow is how we achieve flexible process automation and the workflow engine in the Precisely Automate platform has been specifically developed for complex, data-intensive SAP-centric processes Connectivity through APIs enables you to automate complex processes where your SAP ERP system is the heart of a process that includes coordinating data across multiple systems of record, or where the data needs further enhancing from third-party sources Self-Service capabilities help further enhance automated processes by allowing outside parties to contribute to the process. A great example of this are portals that allow constituents, such as vendors, to access, enhance or modify their SAP data through a secure web page. This can kick off a workflow to review and approve the data, or directly post data to your system. Data Management automate and streamlines mass-data processes Cloud Deployment to support your company’s move to the cloud activities and better connect to multiple SaaS platforms that are engaged in SAP-centric processes. Governance often presents one of the greatest challenges to manual data management processes and to automated processes that rely on multiple technologies to automate a single process. By adopting a centralized approach to automation, you can achieve end-to-end governance for improved visibility and auditability of complex, data-intensive processes Low-code development which is often referred to as “citizen development” enables business teams to develop and deploy automation solutions, relieving much of the burden from our often overworked IT teams. And, by moving the automation development process closer to the people who own the processes and data, it can significantly lower the time it takes to deploy automation across your business. Attended & unattended automation recognizes that while there are processes we can automate that do not require human intervention (unattended), many of the complex processes that are core to the success of our business are “human in the loop” meaning that people are required to input, review or modify data. Having the flexibility to develop solutions for both attended and unattended scenarios can give more flexibility than solutions centered on unattended processes. Forms are perfect for complex processes that require input from multiple parties across your organization. For example, creating a new vendor in your SAP system, often requires data or review from people in multiple departments either in parallel or sequentially. By deploying role-based forms, you are just requesting or displaying the information that an individual needs to see at a [particular stage of the process. The ability to quickly develop and deploy forms in an automated process can improves efficiency in multiple ways. Data Quality - Finally we get to improved data quality. We talked earlier about the interdependence of process and data in achieving automation success. With data stewardship integrated throughout a complex automated process, the Precisely Automate platform offers a more comprehensive set of data quality tools than automation tools that are only focused on process automation. As a complete set of capabilities, the Precisely Automate platform is the right combination of capabilities on which to center your Hyperautomation strategy and build for success.
• #27: Mention citizen development and our history of enabling business teams to develop and deploy automation
• #28: 9 months to realize full return on investment in Studio from improved accounting efficiency (Global logistics company) 1.4 million pricing-change records processed – no additional headcount needed (Home Meridian) (Note: Cannot use Home Meridian name in marketing or promotional materials) Replaced paper forms with online forms for processing customer refunds (RSA Insurance) (Note: Cannot use RSA name in marketing or promotional materials) 300 Journal entries per hour using Excel Workflow (McKesson) (Note: Cannot use McKesson name in marketing or promotional materials)
• #31: Agility to quickly adapt to changes in your business environment through flexible workflow automation Also – Low code/no code environment enables faster solution development through citizen development Speed for faster, more efficient processes Improved initial data quality through elimination of manual data entry and integrated data stewardship capabilities Visibility throughout a complex process for: Reporting Process Intelligence Process linking - real-time process tracking (Does this get called out separately?) Audit support Historical comparison/analysis Business team focus – Low code/no code environment enables business teams and those closest to a process to develop automation solutions Automation flexibility - to determine best automation route (attended vs unattended, form or Excel interface, connection to other systems of record through APIs for end-to-end process automation) Scalability to process almost any amount of SAP data as needed by a process step Proven track record – Nearly twenty years of SAP ERP process automation success has demonstrated why Evolve is the leading process automation platform/solution for thousands of customers world-wide.
• #32: EDM Panel discussion     For the panel: I talked about the core needs (agility, speed and data integrity) and wants (staff optimization, faster ROI, achieving digital transformation goals) that we see coming out of the pandemic, do you have anything to add based on your experience working with customers on a daily basis?   Tim: Yes process and data go together and need to considered two sides of the same coin regarding your digital transformation goals and focusing on one area without the other isn’t likely result in the kinds of faster ROI that require both data integrity and speed. This is why we see many customers using human in the loop automation solution to review and cleanse their data for key use case. Additionally we see many customers centered on fixing their master data first before expanding to their reference data as well as organizing and prioritizing their projects around center data domains and use cases like material master creation, or customer creation and change management, or vendor onboarding.   Martin: I agree with Tim and I would like to add something about the sequence to achieve some of these goals. Without clean master data, process automation for transactional processes that rely on it, or new technologies like AIML that rely on clean master data will not deliver the desired results. And that’s why like Tim said we are seeing many customers understand this and focus on fixing their master data and master data processes first. But sometimes starting backwards, with the transactional business processes and then identifying the master data needs or bad master data can help, for example using process mining, as long as the results are then used to improve the master data and implement processes to keep the master data clean. This introduces an extra step before the transactional business process can be improved and automated and can take some extra time, which means that visible and measurable results for the business processes can not be achieved as fast as sometimes organizations want to see them. But it can be an important step if it is done right, otherwise the overall results can be limited.   Andrew: Segway to AI question   Andrew: So I guess this reinforces a lot of what I talked about with regards to the benefits of artificial intelligence for automation.     AI/ML for Automation Martin: You have been working most closely with customers on our AI POC. Can you talk a little more about what customers are hoping to accomplish/achieve by enhancing the automation of their SAP processes with Artificial Intelligence? Martin: Specifically in our POC, there are two areas where customers are looking at achieving value from AI. One is for their end users in business processes to improve the User Experience and the quality of data. You talked about the interdependence of data and process before, and since AI is a technology that is based on data, AI can identify rules, patterns and logic within data that sometimes are not even visible to users. If this can be brought into a process to simplify the data entry, reduce the risk of choosing the wrong value or to speed up the process by prepopulating fields, then it adds value to the process. The easier you can make a process for the user, and the easier you can make the user enter the right data in that process, the more efficient and successful the process will be.   The other is more for the solution developers and process experts that are needed as part of the creation of these process automations to reduce the complexity and effort in creating solutions, because a lot of the effort is in defining, implementing and updating business rules for example that can be replaced by AI.   Tim: Can you talk a little more about the other aspects of AI that we are investigating at Precisely? How will generative AI play into the future of automation?   We strongly endorse the concept of citizen developers and the transformative power to businesses that enabling the data stewards and the process subject matter experts to be able to own their data and process automations. We believe that AI will play a central role in empowering these business analysis and users to create, cleanse, and manage their business data via automations that they can create and own with the appropriate IT governance and empowerment of their automation. For instance, business users with an excel file filled with thousands of rows of data should be able to leverage AI in their daily automated work to bridge the technical hurdle associated with all the required data APIs of different systems of record like Salesforce, SAP S/4HANA, and ServiceNow. As an end user, I should be able to identify a system of record, a set of data and have the AI service create or identify the existing automation workflow to launch or an automation script to query or load the data correctly into that system of record. Regarding how generative AI will play in the future of automation, I would say that today generative AI’s core capabilities of content creation, information discovery and summarization lend themselves to automation capabilities. The obvious starting point is using a conversational interface to generate data queries that can be automated. Then we should be able to apply associative reasoning to data sets to do things that leverage other forms of AI like predictive machine learning algorithms to populate data fields and this will result in powerful new ways for business end users to find data, transform data, discover data insights and increase a business’ agility due to more rapid decision making and changes to market conditions.     Drilling down a little more on AI and clean data, Martin: SAP is talking a lot about keeping a clean core in SAP S/4HANA. What is your opinion on AI in the context of a clean core? I think AI can add value from two sides. On one hand a clean core can help to augment business processes with AI because as the name suggests the processes and data should be more standard and cleaner. If you manage to have a clean system with clean data, you can build better AI supported processes. On the other I think AI can help keep the core clean and reduce the need to build some of the customized checks etc into the core system. The other advantage in that context that I see is that most of the times implementing a clean core strategy goes hand in hand with standardization of processes and consolidation of ERP systems. Where you might have had 50 or 60 SAP ECC systems before, now customers are consolidating these 50-60 into maybe just one S/4HANA system. But that also means they must consolidate data and processes, which means a redesign and more standardization. And that is creating a better foundation for augmenting these processes with AI.   Fiori automation Regarding SAP S/4HANA, we didn’t touch on this in the presentation, but can you talk about the interest in automating SAP Fiori processes? My understanding is that Fiori is supposed to eliminate much of the need for automation in these SAP processes. Martin: In those apps that are Transactional, meaning used to enter data into SAP, like creating or changing objects, the user still has to enter the data manually. And it is the complexity and volume of fields that determine how complex and time consuming this activity is. A better Fiori UI can help, but it doesn’t eliminate the manual work. And if certain activities need to be done often, and the user would need to perform the task for each record individually, then the need for automation is just the same as with classic SAPGUI transactions. And so the underlying reasons for automation for Fiori apps are often similar to traditional transactions in SAP. Another reason is that Fiori apps are sometimes just used for specific use cases or LOBs and are not the only UI, but coexist with the classic SAPGUI in many organizations. So, very often I see the need for an automation solution that can cover both UIs, the SAPGUI transactions and Fiori apps in a consistent way with a consistent User Experience for Automation.   Can you talk about some unique challenges to automating processes based on true Fiori technology? Martin: One of the challenges with automating Fiori based processes that I see is performance when it comes to mass data use cases. The performance is often not the same as with an RFC based communication. If you need to create or update a large volume of data in a short period of time, I think this can be potentially challenging via Fiori apps. Final Thoughts???   Two quotes come to mind. May you live in interesting times. The pace of change has never been this fast, yet it will never be this slow again. Its easy to become paralized due to the rapid changes that are coming with AI/ML, but its worth considering starting with a small project that will help you deepen your understanding of this technology and how it can be applied in your business automation.   Martin: Automation and process optimization are continuous improvement activities and there is no necessarily an end state where you can say this is the final ROI and we are done. The business is constantly changing and there will always be new technology that can help you increase productivity and automation maturity. In my mind, the only constant here is to establish a strategy and organizational structure that embraces these changes and actively pursues opportunities to improve business processes and to get started.     Seed Questions: What is the timeframe you anticipate that we start seeing AI/ML services adding significant value to automation platforms? Martin: I think we are already starting to see value, as AI is already embedded in many platforms. If it is embedded it might not be visible how much value AI delivers in the background, but I would say it already adds value today for many processes. Do you have a recommended way to measure the value back to the business of automation? Martin: I don’t have a recommend way how to measure per se, because the answer depends on the goals of the organization and the process itself. Whether it is time saved, faster processing time or cost savings, most automation results are measured by quantifiable results. So, if you want to measure the value, you need to have the data already for the current process before automation meaning you have to have clearly defined, measurable and actually measured KPIs. Maybe starting with those that are already established can be a good starting point. How do we figure out where to start and what processes are the best ones to automate? Martin: I find that in many organizations the manual processes and inefficient processes are known to a certain extent. And tools like Process Mining can help identify these processes. In my view, it is typically a good starting point to look at the core business processes first. What will AI/ML do for citizen developers? Martin: Essentially citizen developers need a layer between the actual code and the development UI for them. The between this abstraction layer is and the easier it makes it for Citizen Developers to using building blocks to build solutions the better. I think AI has the potential to create these building blocks easier and better and provide the Citizen Developer with more powerful components to use. Can you give some examples of using APIs to enhance process automation? Martin: Augmenting new product creation with AI for Data Attribute Recommendations, Updating PSP elements in SAP out of MS Dynamics, Updating Vendor accounts from Salesforce to SAP, Our company is still trying to figure out our automation strategy. Do you have any tips on how to get started? Martin: If you look at the goals of Hyperautomation for example which are to automate as much as possible in a concerted way, it has these two components of 1.) identifying how many processes to automate and 2.) how to bring automation solutions together in a concerted way to accomplish the automation of these processes in the best and most efficient way. In my mind, it starts with identifying and understanding the automation potential and then building out the strategy from there.
• #33: Seed Questions: What is the timeframe you anticipate before we start seeing generative AI offered as an automation platform? Do you have a recommended way to measure the value back to the business of automation? How do we figure out where to start and what processes are the best ones to automate? What will AI/ML do for citizen developers? Can you give some examples of using APIs to enhance process automation? Our company is still trying to figure out our automation strategy. Do you have any tips on how to get started?