This is the write-up from the tenth in our series of virtual coffee breaks – designed to provide a caffeine-enriched forum for discussion amongst senior executives who might otherwise be going stir-crazy during this period of lockdown.
The topic for this session was: ‘Can IA rescue disrupted businesses and deliver complex end-to-end processes? We look beyond the limitations of RPA to find the answers’ and our speaker was Richie Barter, founder and CEO of AltViz.
AltViz help companies improve business performance by allowing them to access, analyse and act on data, and automate and scale critical business processes, using their Integrated Intelligent Automation Platform.
Richie opened the conversation by welcoming everyone and highlighting three important things for any company, that has invested in Robotic Process Automation (RPA), and is now thinking about moving to the next level, which is Intelligent Automation (IA):
RPA has been a hot topic in Enterprise Architecture for around ten years. Essentially, it’s a way of automating tasks that would otherwise be carried out by “fingers”. In other words, tasks that involve someone typing at a keyboard.
The reality is not all RPA initiatives have delivered value. In fact, Gartner estimate around 50% of RPA projects have not delivered the business benefits they set out to achieve. Companies are also finding that, as the complexity of automation increases, RPA further loses efficacy.
Pushing the analogy to the next level, IA is a way of automating the “eyes, ears and brain” using optical recognition, natural language processing and artificial intelligence and, importantly, a “nervous system” to tie all this together (which is where AltViz’s Integrated Intelligent Automation Platform comes into play).
As automation leaders start to look at the implications of digital transformation, they’re realising the shortcomings of RPA and asking whether IA is the answer – IA uses more complicated algorithms, handles both structured and unstructured data, and drives downstream processes, via API’s and automated messages.
IA also provides greater flexibility in adapting automation in response to changing circumstances, as highlighted by the current Covid-19 situation. Some of Richie’s customers have found the current crisis is highlighting gaps in their processes, for example:
In both cases, IA can help to enable these new ways of working without having to rip out the existing RPA or make extensive changes to the underlying core systems.
In terms of automation, many organisations have tackled the low-hanging fruit – individual use-cases in specific functions or business areas. As they look to scale this automation and start to think about coordinating their efforts to maximise their returns from this investment, this is where IA makes a difference.
Richie’s clients will usually talk through their use-cases in terms of outcomes and tangible business benefits. The use-cases where IA can have the greatest impact are those where there is a defined business benefit, a complex end-to-end process, multiple disparate data sources (some of which may be unstructured – e.g. voice recordings), complex decision logic and multiple actions that flow from this.
A couple of examples from AltViz clients in different sectors:
By working with a well-tuned team and an easily configurable IA platform, the value of IA can often be demonstrated in just a few weeks, with the IA plugging into core systems via API’s and providing automation that is safe, resilient and auditable.
The move to cloud-based solutions is further accelerating IA, because it makes connecting to data sources much easier. You don’t need to write lots of different API’s to disparate systems, just one connector to the cloud data store.
Richie finds that customers, who often think in terms of six-month windows for implementations (having previously been burnt), are surprised to see a working version of the IA up and running in as little seven days.
Use-cases that don’t lend themselves to IA are generally those where there is a low volume of data or unique, one-off processes. For example, the area of catastrophe insurance has a lot of one-off, long-tail events. It’s hard to build IA around these sort of use-cases.
Another area that is not really suitable for IA is where you haven’t got good data – the rule of rubbish in, rubbish out, applies to IA just as it does to any form of automation. Having said that, it’s not always a deal-breaker. For example, when manually reviewing voice calls you might only be able to cover 1% of calls, and even though voice-to-text won’t get you to 100%, it will get you far enough to make a difference.
If you would like to learn more about the Hovis and Greenergy case-studies that Richie referenced, please visit:
If you would like to learn more about our other events, please click here.