The Future of ERP Software Is Self-Driving

Posted by  Predrag (PJ) Jakovljevic

Gone are the days of enterprise resource planning (ERP) software systems as unwieldy repositories of transactions and records. Even the savviest power ERP users had to know where to look for the information they needed to find. Any custom reports they developed and analyses they performed could only tell them that something had already happened, and perhaps why it happened.

ERP software users have traditionally spent too much time verifying the past. This involved time-consuming, tedious, and menial tasks. Think of batch processing, fiscal period closing, periodic audits, timesheets, purchase requisitions, travel and expense (T&E) reports, time and attendance (T&A) management, human resources (HR) benefits enrollment and administration, invoice management, opportunity and risk assessment, and more.  

The days of number crunching, data entry, email and calendar management, and similar repetitive tasks may be over. Although these activities help to identify issues, it’s typically too late to act on them.

Enter Smart ERP

Today, smart ERP software systems are being developed using artificial intelligence (AI) and machine learning for cognitive capabilities, striving to understand context and to anticipate the users’ needs. They should be able to discern what might happen next, alert users to that activity, and even prescribe some advice on what to do about it.

Given that manual tasks are largely an exercise in pattern recognition, machine learning algorithms should be able to quickly and efficiently mine volumes of documents and transactions. Intelligent automation has the potential to discover hidden patterns that humans cannot easily discern buried deep within bronto-, yotta-, zettabytes of scattered and disparate data.

New ERP software systems are increasingly being built on flexible cloud architecture designed from scratch on microservices, faster processors (with in-memory data management capabilities), and open application programming interfaces (APIs) for easy integrations and extensions. Increasingly, they also have built-in AI and workflow engines.

Thanks to the existence of a number of commercially available cognitive services by Microsoft, Amazon, Google, and even open source software communities, ERP software providers don’t have to reinvent the AI wheel. Microsoft, for example, offers several cognitive AI services, such as Office Delve (for Office 365), Microsoft Azure LUIS (Language Understanding Intelligent Service), Face API, and the Microsoft Bot Framework and Azure Bot Service to support natural language processing (NLP), machine learning, deep learning (face recognition), and other intelligent tasks.

Digital Assistants for ERP “Driving”

Thus, some ERP software vendors have recently come up with their own versions of digital assistants like Siri, Alexa, and Cortana. Think of Unit4’s Wanda. Other vendors have their own catchy chatbot names.

These digital tools can interact with the user in multiple ways, via voice or text, through Skype, Slack, and Facebook Messenger, diminishing our reliance on cumbersome and data field-intensive ERP user interfaces (UIs). What’s more, via the face recognition capabilities we expect these bots to have soon, the user will be able to avoid tedious system authentication log-in processes (who remembers all their passwords?).

For example, behind the curtain and always on, these bots could spot irregular patterns or incorrect transactions, alert users to events, and prompt them to review certain issues. The digital assistant will also increase ERP user productivity, starting with pre-populated forms and in-context yes-no validation questions pertinent to each persona.

By monitoring and learning from users’ ERP software usage, these bots can continually adjust what users see on their screens to reflect user preferences and behaviors. As these AI tools become smarter, they will learn from the past and actively make use of the big data and knowledge that is captured within the ERP systems to provide context-sensitive support for important tasks. People will have the data and the time to make real strategic and value-adding decisions for their companies.

The Ultimate Goal: Self-Driving ERP

The marketing message from ERP vendors like Unit4 promotes “self-driving ERP” software, but today’s ERP software still needs a little more human input than just a final destination. In the case of ERP, we are still talking about “augmented intelligence” more than AI in its true sense.

Bots can currently perform the ERP “driving” tasks like cruise control, self-parking, lane-keeping, automatic braking, etc., but not true autonomous driving. For now, the idea is to use intelligent software tools to do the grunt work (under guardrails), with human experience, intuition, and know-how being critical to supervise and manage by exception. But the near self-driving ERP utopia is perhaps not that far in the future.

Predrag (PJ) Jakovljevic

Predrag (PJ) Jakovljevic is a senior analyst at TEC, a global consulting and advisory firm helping organizations select the best enterprise software solution for their needs. He has over 20 years of industrial experience within the discrete manufacturing sector, including the machinery and equipment, automotive, construction and engineering, and electronics industries.