AI-driven development with OutSystems Mentor
I founded my first company in 1997 and soon realized that software development projects were never delivered on time and on budget. Much of the difficulties came from the burdens of traditional code: crafting it by hand was a very time-consuming activity reserved for elite professionals who could easily be needed in a different project, or simply quit their job, leaving others to understand what they had coded.
Additionally, traditional waterfall methodologies used to fail because the business changed during the project, forcing a redesign of the architecture that crossed over databases, business rules, front ends, and integrations. Adapting the dependencies of all that code further delayed the projects. 2+ year projects became 3 year projects, 3 year projects got further extended. CIOs had to face the hard reality that a lot of big projects either got canceled or were delivered below business expectations already carrying a large backlog of highly expensive changes.
That's why we focused on the anatomy of change. We started looking for ways to accelerate change, by automating code dependencies and removing common pitfalls we found in projects. We spent a lot of time mapping the patterns we could automate and defining models that we could replicate to help IT move faster.
The result was the birth of OutSystems, a visual development and operation platform that fundamentally accelerates change by infusing automation into the entire software development lifecycle. And 13 years later, the analysts came up with a name for what we did: low-code.
The GenAI promise
Fast forward to 2022. While automation was already a part of the daily work of many developers, low-code developers in particular, the GenAI disruption promised to accelerate work for every developer. We witnessed the quick adoption of GenAI-powered "copilots" that enhance the productivity of teams by helping them generate code faster. Two years later, GenAI has become part of the software development process across the industry.
However, if left ungoverned, generative AI will also accelerate risks. And once we started running benchmarks around GenAI being used as a coding copilot, we began to realize the trade-offs of using these tools—because there really are no free lunches in this industry.
No free lunches
One issue is that the quality of code has become increasingly harder to control. Copilots enhance team productivity by helping developers generate code faster, but they're also a driver of technical debt. AI-generated lines of code, besides being inherently unpredictable due to the nondeterministic nature of AI models, tend to be longer than what you'd need if you wrote them by hand. So, as you read this, hordes of bots are generating millions of lines of code that are human-orphan at birth: not all developers will fully understand what was generated (or why) and will end up needing another tool to make sense of the code the first one generated.
Also, AI-powered coding assistants only address code, which is a very small part of a software development project. That means you’ll need different tools to ensure critical aspects like explainability, governance, and the security rules and policies that keep generated code under control.
This is why we believe that, ultimately, code will become buried underneath the interfaces that can provide more visibility and control into the full software development lifecycle (SDLC). And this was the vision that guided us in the creation of OutSystems Mentor, an AI-powered SDLC digital worker, on OutSystems Developer Cloud.
Mentor is the natural evolution of what we had in mind when we created OutSystems: to build a platform for rapid, high-quality change. Offering increased agility–now, AI-powered. Mentor automates critical tasks and can transform natural language or product requirements documentation into full-fledged applications that are immediately usable, easy to maintain, and fast to evolve.
Low-code x AI = infinity
With Mentor, OutSystems become the first low-code platform to help enterprises harness the power of GenAI across the Generative Software Cycle without wreaking havoc. Mentor supports teams in the development stages, from app generation to optimizing the building, testing, and deployment stages, and assists operations by keeping applications secure, highly performant, and maintainable in the long term.
This means faster, more consistent, and more scalable development cycles, with fewer errors. No surplus code, no orphan code, no security or compliance deviations. Mentor will fundamentally make you immune to tech debt.
We firmly believe that the future will bring an acceleration of integrated AI-powered platforms like OutSystems, which combine visual development with natural language to compress the time it takes to build, maintain, and evolve applications.
The combined power of low-code and AI unlocks infinite new possibilities for organizations, dramatically enhancing development productivity and delivering unique software capabilities. As the Generative Software Cycle evolves, I’m excited about how we’ll evolve OutSystems further to reduce the toil and compress the time of development by orders of magnitude.
