How Enterprises Can Combine Decision Automation, Machine Learning And Process Automation For Better Business Results – Forbes

Every day we hear about new ways automation is transforming businesses and customer-facing applications – for example, insurance claims processing or automated retail support services. It’s no surprise recent estimates forecast enterprise AI spending is expected to reach $500 billion by 2024. While AI holds the potential to transform all businesses across industries including financial services, insurance, retail, healthcare and more, many organizations find it challenging to build a solid foundation for scalable change. 

To further understand the differences in automation technologies and their role in enterprise transformation, I recently spoke with Alan Young, chief product officer at InRule Technology, an AI-enabled, end-to-end automation provider.

Gary Drenik: What is the difference between decision automation, machine learning and digital process automation? How do they all work together?

Alan Young: Decision automation is the act of automating the consistent application of business rules and logic governing an enterprise operation and behavior. Developed by business experts, decision automation can help establish facts, identify patterns, make choices, trigger processes, determine compliance and surface knowledge. Through decision automation software, both technical and non-technical stakeholders (data scientists, developers, business decision-makers, etc.) can define, share, and execute decisions in real-time and embed them within applications. This allows organizations to make real-time, repeatable, and complex operational and customer engagement decisions in a scalable manner. Further, the introduction of machine learning offers the ability to apply non-declarative, probabilistic decisioning options and logic. To keep up with evolving market conditions and competition, this enables modern businesses to act on real-time, dynamic data inputs and values.

Digital process automation (DPA) helps manage and orchestrate business processes and workflows within an organization that implement a business operation. However, DPA does not just handle simple tasks within its reach. In fact, the large majority of enterprise workflows involve highly complex and extensive tasks, such as processing retail purchases, that involve the interoperation of data and activity exchange among any number of enterprise applications and services orchestrated by the DPA tool. In this context, DPA relies on decision automation platforms to help apply business rules and policies consistently.

In practice, all these technologies play a role in automating enterprise functions that have evolved from operating on static, pre-gathered data to become highly dynamic and contextual. According to a recent Prosper Insights & Analytics survey, 61% of US Adults age 18+ said that low prices are very important to them when shopping online. Therefore, pricing needs to be dynamic for products, based on user location, product availability, tax, shipping, and more. An integrated approach to automation can help power this operation. In practice, decision automation can help calculate retailer costs, machine learning can predict the optimal price that will lead to purchases from buyers, and process automation can expedite fulfillment and shipping.


Drenik: Why do enterprises need all three solutions to be successful?

Young: For enterprises, an end-to-end approach is critical for successful business automation. By integrating these three solutions, enterprises can achieve scalability to execute processes time and time again, making organizations faster and more agile, and with a greater competitive capability to adapt to threats and opportunities in minutes or seconds rather than days, weeks, months, or years. In increasingly competitive marketplaces, scalability and speed are critical differentiators.

Today, we live in a digital-first world, where consumers expect efficiency, convenience, and engaging brand experiences. Prosper Insights & Analytics data also shows that over 50% of US Adults 18+ say that website ease of use is very important when shopping online. Machine learning can be leveraged to ensure personalized experiences and product recommendations, but decision and process automation that implements a retailer’s business goals, policies, and strategies for product and pricing are still needed to effectively drive and leverage machine learning.

The same Prosper Insights & Analytics survey also shows that about 22% of Gen-Z and Millennial shoppers are influenced to purchase products based on advertisements on Facebook or Instagram. To provide successful recommendations, retailers can leverage machine learning models to tailor advertisements based on previous buying behaviors and user preferences. However, an integrated approach with decision and process automation is also needed to ensure these recommendations are scalable and based on real-time inventory and pricing.

Drenik: How can enterprises get started if they don’t have any of these capabilities or if they have just one?

Young: For enterprises considering automation initiatives, the first priority will be to establish a single source of truth to inform future business logic and rules. Decision automation technology offers a productive foundation to create a centralization of business rules, logic, and goals. This approach helps improve accuracy for future machine learning models and can accelerate ongoing initiatives to operationalize AI. 

Decision technology also allows organizations to extend capabilities of existing applications, deliver new services, provide auditability for regulated industries, and efficiently adapt to changing market conditions. Businesses make hundreds and thousands of decisions every day, and decision automation can help improve operations and execute decisions with maximum flexibility and scale.

Drenik: How can enterprises unlock automation’s full potential?

Young: Digital transformation is still the key priority as it relates to business modernization. Organizations need to rethink their legacy systems to best compete and comply with dynamic regulations in today’s market. Low-code and no-code platforms will increasingly allow organizations to build applications and automate decisions, processes, and more. In addition to embracing low-code technology, organizations must still solve two key challenges to unlock the full potential of automation.

Recent research shows more than half of organizations feel they have too much data to make collaboration efficient, creating roadblocks to AI project success. While data plays a key role in successful automation, organizations must first master data in motion to efficiently capture, configure, process, and maintain data in real-time, regardless of location in the tech stack. Second, automation can create a “black box” challenge for companies. For example, the above research also shows 58% of decision-makers find defending or proving the efficacy of their digital decisions challenging. As automation takes over more mission critical processes there must be a dedicated effort to improve explainability and transparency among AI and machine learning processes. In doing so, organizations will have a mature and scalable automation strategy to support long-term business development and operations.  

These digital transformation efforts will also introduce a new universe of human-centric workflows that depend on decisions, traditional processes, and machine learning to bring richer digital experiencies to end-users and consumers. Based on precise, real-time data, these efforts will lead to more effective and efficient business outcomes.

Drenik: What is the key to democratizing AI for all enterprise employees?

Young: The primary challenge for successful democratization of AI lies in breaking down internal silos and creating clear communication channels. Collaboration among all business stakeholders is vital. However, research shows that one in four organizations do not embrace a culture that encourages data democratization, which further exacerbates these internal silos.

By embracing technologies that require little or no programming effort, organizations can empower all employees – domain business experts, data scientists and delivery teams – to inform and integrate predictive logic into business processes for successful automation. This collaborative approach also allows for more efficient and agile innovation.

Drenik: How do you see enterprises using automation in 5 years?

Young: Automation will become more and more accessible to enterprises of all sizes and industries. The ability to innovate around automation will open up multiple dimensions of competitiveness, including user experience and contextual data utilization leading to profitable business outcomes. Automation will help improve customer satisfaction and experiences, and achieve smart, data-driven decision making with less effort. As organizations automate more mission critical processes, humans will still have an important role in business success. For example, in retail, automation can help support customer-facing applications, order processing and inventory management. However, employees will still handle the tasks that require further oversight and more intelligent thinking. Supported by intelligent data extracted from automation insights, humans can improve their own skills and provide greater impact in areas including one-on-one customer support or business planning. This approach to human-augmented intelligence will help achieve greater business productivity and profitability.

Drenik: Thanks Alan, we appreciate your insights on the future of enterprise automation and the importance of transparent, data-driven decision making. As organizations continue on their automation journeys, it will be exciting to see how these technologies further transform customer experience, daily life, and workplace roles.

Spread the love

Leave a Reply

Your email address will not be published.