How AI might make ERP friendlier

Published on the 06/06/2017 | Written by Donovan Jackson

Friendly ERP systems

Anything that makes logging into your ERP a happier moment has to be a good thing…

It’s probably not a good idea to start a conversation with someone who delivers & supports ERP systems by saying ‘ERP systems are not generally nice things’.

But that’s where we begin with Liz Pay, senior consultant at Agilyx, which provides and supports Unit4 software in Australia and New Zealand. Perhaps an odd point on which to kick off, but the topic of the day was, after all, how ERP systems are evolving to become better not only from a functional point of view, but a usability one.

“Over the last three or four years, these systems have been getting friendlier and that is part of a development continuum,” responded Pay. “There is a lot of focus on changing the centre of ERP from being a financial transaction to being a user experience and interaction.”

That started with, for example, mobility options, putting ERP-type functions onto the smartphone to do any one of a variety of (typically) admin-type tasks in the micro-moments on the train or bus or what have you. “That’s a step in the right direction, but it isn’t revolutionary. It still requires the user to decide what they want to do. It might simplify data entry, but the process is essentially the same.”

With artificial intelligence all the rage (again) and chatbots going from peculiarity to popularity in what seems like a matter of months, the tantalising possibility of ‘sensing’ ERP systems has hove into view. Unit4, which already claims to be among the friendlier of the many ERPs available today, recently introduced its variation on the theme. In keeping with trend, the chatbot has a disarmingly nice name: Wanda.

But could this development and others like it signal another wonder of the tech world? Pay reckoned that yes, this could push that continuum along a good way. “What we are finding with the Wandas is that they are addressing the ‘assist’ part of ‘assistant’. Just like an assistant, the idea is that they are predictive and prompt you to do the things you would normally do at any given time of the day. So, it will remind you, for example, to fill in timesheets at the appropriate time of the day. It will analyse your daily routines and actions and make suggestions and prepopulate, rather than waiting for you to remember what you need to do.”

Sounds great, because for most of us, it is a major challenge to remember to, or get around to, doing the ‘admin’ things which, while important, just aren’t fun. “Right now, the focus is on the ‘hygiene’ factor, but the potential for more is there. For example, soon you’ll be able to just ask the assistant for a report and it will know what you mean and where the data is. Now that’s the ‘assist’ part of an assistant,” Pay added.

As the traditionally clunky world of ERP is further imbued with AI and helpful chatbots, there is every possibility, agreed Pay, that the necessity for change management as a component of any major systems adjustment (whether a new implementation or a significant upgrade) will diminish. “You find that already with intuitive systems which are designed to be more obvious to use and which are consistent with the user interfaces of other [often consumer] applications. That does mean you need less training; when we get to the point of being able to ask a literal question of the system in natural language, and it can provide answers specific to your industry and business, well, that will be hugely beneficial.”

But Pay stressed that such functionality cannot come at no cost, unlike a lot of the consumer chatbots which are built into smartphones. That’s because to be really useful, the bots will have to be taught using specific data sets – otherwise, the ‘help’ will be so general as to be useless. “It won’t ‘just happen’. There has to be an investment in the background; this is supervised machine learning and that does take some effort.”

Just how far is the sort of help represented by Wanda from being ‘in market’ in Australia and New Zealand? Right now, it’s on the horizon. “It is reasonably new and we don’t have a working version yet, but we are expecting that end June. But the promo videos demonstrate the potential and our clients are excited by the possibilities.”

Pay is sanguine about the immediate prospects of ERP chatbots, but she added a good dose of caution over inflated expectations, too. “We’ve all seen Terminator and with that tend to have excessive expectations of what AI can do. But those expectations do give us something to work towards.”

And that keeps the momentum going along that continuum just fine.

Post a comment or question...

Your email address will not be published.

Time limit is exhausted. Please reload CAPTCHA.

Follow iStart to keep up to date with the latest news and views...