OutSystems may be best known for its low-code development platform expertise. But the company has steadily been moving to a specialism in AI-assisted software development – and the parallels between the two are evident.
In June, the company unveiled its generative AI roadmap, codenamed ‘Project Morpheus,’ with benefits including instant app generation using conversational prompts and an AI-powered app editor offering suggestions across the stack. The mission remains clear: ‘developer productivity without trade-offs’, as founder and CEO Paulo Rosado puts it.
Project Morpheus, in the words of Nuno Carneiro, OutSystems AI product manager, is ‘the next generation of software development.’ “What we’re doing is building a completely new development experience, based on this premise that AI will give you suggestions. You do not have to code practically anything, and the AI is suggesting what to do,” says Carneiro.
“You have a What-You-See-Is-What-You-Get visual experience in terms of software development where you can change the application directly in your development environment. – On top of that, AI gives you suggestions about what you might want to change so that you don’t need to code things manually.”
This means the artificial intelligence is there to tweak, rather than take over. The company’s main offering in the space to date has been the OutSystems AI Mentor System. From code, to architecture, to performance, the developer is in control, but always has an on-call expert to hand.
Scepticism is naturally there, as it was with the rise of low-code platforms. But having slayed the dragon once before, is the job easier this time? “We see the same patterns of people being sceptical of AI in software development,” explains Carneiro. “We’ve been through this process of educating and showing the value of automation in software development before. We now feel like we’re in a good spot to communicate the current transformation in the industry due to the rise of AI.”
The key factor is that the OutSystems platform guards against some of the less salubrious aspects of artificial intelligence technology. Hallucination – where an AI confidently gives an incorrect response – and creating code riddled with vulnerabilities are just two of the pitfalls which could result if given full control. This is where the parallels between low-code and AI-assisted software development are especially striking; even if the code has been generated by AI, you can visually understand what you are building.
“The solutions we see out there at the moment still don’t solve this problem,” says Carneiro. “Because if AI is just writing a bunch of code automatically, and the person in charge of seeing the code and building it doesn’t understand what’s behind it, that’s not going to be a solution for any serious organisation to use. Low-code solves this problem with its visual development experience and the AI Mentor System constantly checks for security vulnerabilities, no matter who, or what, wrote the code.”
The bottom line for businesses is that AI-based development with a low code platform will allow them to complete projects in weeks which would otherwise take months, or even years, to develop. Carneiro gives a theoretical example of a company who wants to do a proof of concept for a new piece of software managing HR internally; a project which could take a week with OutSystems. For wider transformational projects, such as rebuilding an entire supply chain, it would take a few months maximum.
There is another benefit too for larger firms. “We’ve also seen a lot of clients build Centres of Excellence around low-code software development that they then export to their organisations around the world,” says Carneiro. “Using the AI Mentor System means they can then export this and innovate quickly across their whole business.”
Improving the process of software development is only one aspect of a digital transformation journey, however, with OutSystems committed to enabling businesses to adopt AI themselves. Image recognition is one such use case, or using cognitive services that users can add to their applications to solve business problems from unstructured data. This was factored into one part of the generative AI roadmap update, with a new connector announced for Azure OpenAI, built in partnership with Microsoft, to enable the use of large language models in development. “Part of our roadmap here is to help customers build the foundations for AI adoption in their businesses, so they’re not caught off guard,” notes Carneiro.
OutSystems is participating at AI & Big Data Expo Europe, in Amsterdam on September 26-27, and AI and wider digital transformation journeys will be a major part of the agenda. “A typical digital transformation challenge is to connect different data sources, and that’s another place where we believe OutSystems comes in. We’re at the right spot to help businesses solve this,” explains Carneiro. “We naturally help you connect with different data sources, and it’s something we’ve been optimising over the years to help our customers bring in all types of databases and sources – we have tools that help customers connect to integrations and integrate different data sources.
“These challenges might not be obvious before you embark on an AI adoption journey,” Carneiro adds. “But I’m pretty sure anyone who’s tried will recognise them – and we hope they also recognise that OutSystems is a good partner for that.”
Looking to revamp your intelligent automation strategy? Learn more about the Intelligent Automation Event & Conference, to discover the latest insights surrounding unbiased algorithyms, future trends, RPA, Cognitive Automation and more!
Explore other upcoming enterprise technology events and webinars powered by TechForge here.