Unleash twice the smarts to uplift your processes

As Australian businesses address expectations of operating in a digital-first world, they are under considerable pressure to bring their business processes up to what end users now consider ‘par’. Many will turn to automation to close the gap.

Process automation can take many forms.

Some automations are simple file-based data exchanges between systems, or perform optical character recognition, document workflow, and document and record management. The term also covers more complex use cases, such as systems orchestration, where it performs business-critical transactions spanning a myriad of backend systems or involving armies of cloud-interconnected machines.

On the other end of the complexity spectrum, manual, low value, multi-system processes usually performed by one or a very small group of specific individuals are usually automated using Robotic Process Automation (RPA) technology platforms.

One of the challenges with RPA, however, is that when the technology encounters the slightest change – such as an unexpected name in a spreadsheet – it results in an exception that has to be dealt with separately. The process then has to be re-mapped to be able to recognise and handle a similar occurrence in future, which results in time and money wasted.

Because of this impediment, key RPA vendors have introduced improvements in the form of Artificial Intelligence (AI) and Machine Learning (ML) modules that can handle a level of uncertainty. The combination of RPA and AI/ML is called ‘Intelligent Automation’.

Two recent surveys agreed that more than 70% of businesses are uplifting their processes with Intelligent Automation.

You can imagine some processes where this could be useful. For example, in a compliance-related process that captures information on different public website portals, an Intelligent Automation approach could handle the slight user interface variations that occur over time.

Another major advantage of a Machine Learning module is its ability to recognise patterns. This is especially useful when dealing with information extraction from documents where formats can vary, such as a purchase order or an invoice. When it comes to natural language recognition (text to speech or speech to text), AI and ML modules become an indispensable ally.

Though the advent of Intelligent Automation is extremely promising, do not believe everything the vendors say in their marketing brochures. Even though a fair amount of smarts is included in an off-the-shelf RPA package, it will be down to you or your implementation vendor to train the system appropriately.

Make sure you work with people who know what they are doing – and that can deploy Intelligent Automation intelligently.

Doubly smart

It is worth unpacking this further.

First and foremost, customers that go down the Intelligent Automation path need to assure themselves that what they put into production or out into the market actually displays adequate intelligence.

Customer service chatbots are a prime example. As users, when we encounter these bots, their ‘intelligence’ and ability to answer our questions varies considerably.

By default, customer service bots can read a knowledge base and extract a fundamental relationship between basic questions and possible responses.

But to get to the next level of intelligence, a lot of work is necessary to cover the subtlety of what makes a valuable conversation. The bot may need to be trained in industry-specific language, or in natural language variations that a single question might be asked especially if it involves a lot of jargon or acronyms. Then there’s the response the bot produces: does it reflect the culture, tone and ‘voice’ of your organisation? All of these factors can make a customer service bot sound much more intelligent and user-friendly.

It is not enough to add intelligence to a process automation; an intelligent overall approach to the project itself is also required.

Intelligent Automation projects require careful and proactive change management and communication. Bots are added to traditionally human-led or run processes. That needs to be sold both to staff that work alongside the bots and to end-users that will interact with them. A lot of communications and change management is therefore required because of that human aspect – and you can’t expect your RPA tool to do that on your behalf.

Organisations need to be able to reassure staff and users that Intelligent Automation will benefit them.

Working with professionals that understand the technology can help you communicate the introduction of automation and avoid damaging rumours or negative sentiment. Having users onboard and even excited by the changes ahead is fundamental to ensure the success of an automation project.

Engaging with a partner like Tecala means access to proven methodologies and resources to bring new levels of intelligence to your day-to-day business processes and operations.