Low code systems are already being used in conjunction with robotic process automation (RPA) to allow business users to manage automated business process, without the need for specialist IT skills. As process automation becomes infused with AI – a phenomenon known as intelligent automation – low code tools will help those business users automate more sophisticated tasks, and help DevOps teams manage the complexity of modern application development.
Experts predict it will hasten the emergence of ‘business technologists’, who design and implement automations for a specific business context.
The role of low code in process automation
Low code adoption is forging ahead across industries, with 77% of the 2,025 IT professionals surveyed for the 2021 State of Low Code report published by vendor Mendix saying they have already implemented low code solutions within their organisations. In many cases, these companies are using low code platforms in order to automate data gathering and business processes.
Low code systems are ideally suited to allow non-IT staff to manage and modify automated processes, says Eric Tyree, SVP for AI and innovation at UK RPA specialist Blue Prism, which has just been acquired by US fintech company SS&C for £1.2bn. "[Low code] is central to the entire problem of automation," he says. "One aspect of that is the economics - if we want companies to have hundreds of digital workers, or robots, the process of creating and managing automations with these robots needs to be quick and inexpensive. The way you achieve this is for the people who actually set up the digital workers and manage them to be the same people running your operations. You're not relying on technical staff."
The other benefit of low code for automation, Tyree says, is getting staff up to speed quickly on interacting with the automation process. "At the moment interactions between humans and digital workers are an edge use case for many businesses," he says. "If you want that to become central to your business, you need the average office worker to be able to interact with the digital worker, to be able to give it information, receive information back and not need to go to school for six months to figure out how to do this. For that, you need low code tooling."
As well as serving non-IT staff, low code platforms are part of a suite of tools that in-house DevOps teams are using to meet the scale of demand for new applications and functionality, says Charlotte Dunlap, principal analyst for application platforms, enterprise technology and services at GlobalData.
"To reduce risk and lower costs associated with digitisation, operations teams are seeking DevOps tools in the form of newly consolidated platforms," Dunlap says. "The ability to apply low code, AI, business process management (BPM), and robotic process automation (RPA) to the principals of building automated workflows reduces the barriers of moving from monolithic to microservices-based applications and the cumbersome data integrations necessary to create and maintain each microservice."
Vendors such as Mendix and Appian offer low code development platforms, while the cloud hyperscalers also provide similar tools to users through products such as Microsoft's PowerApps suite and Google's App Maker. The market for these tools is expected to reach a value of $65.15bn by 2027, up from $12.85bn in 2020, according to figures from Brand Essence.
Can low code tools deliver true intelligent automation?
Low code tools are already used to allow businesses users to interact with RPA implementations, but can they handle intelligent automation processes that incorporate AI? Yes, says Dunlap. Modern low code tools "are advanced enough to create and leverage AI", she says. "These platforms have been AI-injected for a few years, and this has increased their importance to developers and non-developers involved in the app modernisation process."
The latest advance is the increasing inclusion of machine learning models, which can be trained to carry out more advanced tasks, she says. "We expect more of these to be incorporated into low code solutions," Dunlap explains. "This will help address more complex workflows, especially for businesses in highly regulated industries and those aiming to digitise mission-critical applications."
This will allow more widespread automated "observability", Dunlap predicts. "You will see [low code platforms] with the ability to have more predictive and prescriptive analytics," she adds. "We expect to see this happening over the next year. Observability will get tucked in with that, and we're probably going to see more proactive automation, so if they spot an issue in your network they can address it automatically."
Indeed, the integration of AI will make low code more flexible, Dunlap says. "There's definitely less scope to customise your applications [with low code]," she says. "You don't hear developers waving aside low code because they know the benefits it can bring, but there are limitations, particularly in the back-end integration which professional coders need to attend to.
"I think that's another area we're going with machine learning, which could help professional teams tremendously going forward."
'Business technologists' and the changing role of IT teams
Research house Gartner says the advent of low code systems driving intelligent automation has given rise to a new hybrid generation of workers termed 'business technologists'. Jason Wong, distinguished VP analyst at Gartner, says these people "are not IT employees, but are using tools like low code, intelligent BPM, RPA, and even data science tools, to drive up the level of automation in their organisation".
Last year, Gartner surveyed 12,000 workers across a variety of industries and discovered that 41% fit the business technologist description. "These people fit two broad categories," Wong says. "There are people who are hired to make a particular technology work. These people tend to be from a technology background but have moved into another area, so you could get a Python developer in marketing, or a data scientist in accounting." These people represent about 20% of the business technologists identified in Gartner's research, Wong says.
The other 80% are non-technical people who "in the course of their job or in addition to their job, are creating technology output for themselves, for their teams, departments, and even sometimes for customers," he adds.
For tech leaders, finding ways to fuse the skills of these business technologists and dedicated IT teams will be a key challenge in future as automation increases, Wong says. "More and more companies that are standing up citizen development programs where they kind of encourage the self-service access to tools and building of prototypes by business teams," he says. "But in general we find business technologists are part of multi-disciplinary teams, which marry professional technologists and business domain expertise. Our research shows 80% of these teams report into a line of business, not the IT department."
Dunlap agrees that more cross-functional working is inevitable as the relationship between low code and automation grows. "There's always been a certain element of 'us against them' when it comes to IT and operations," she says. "Now that's being redefined and there's more co-operation, not least because it's becoming harder to get your hands on professional developers, and companies are looking for ways to alleviate the workload of the technical staff they do have. Bringing more people in helps with digitisation and automation, and that's a good thing for businesses, not to mention people's careers."