Is AppGen a threat to SaaS? OutSystems customers react

A recent Gartner study found that 67% of CIOs are making cost optimization a priority for 2025 and beyond. They're rethinking investments made over the past decade, such as legacy systems, duplicate SaaS tools, and monolithic platforms that are hard to update and integrate. However, thanks to new technologies, it is becoming more cost-effective to modernize and streamline these systems, turning outdated platforms into efficient, connected solutions.

At the heart of this shift is AI-driven software development, but not all AI is created equal. AI-powered application generation (AppGen) is accelerating the “build” side of the build vs. buy equation, making it faster than ever to create applications from scratch. Some experts, like Forrester’s John Bratincevic, argue that AppGen poses an existential threat to enterprise SaaS, as AI-generated apps could reduce the need for traditional software vendors.*

Balancing AI-generated apps with agentic AI to create strategic assets

At the same time, AI-generated apps can’t meet all needs. That’s where agentic AI comes in. Agentic AI creates a shift from conventional narrow AI that is programmed for specific tasks. Using large language models, massive training datasets, scalable computing power, and connectivity, agentic AI can set its own goals, plan optimized workflows, make nuanced decisions, and adapt to changing circumstances. This autonomy unlocks new frontiers for automating complex, end-to-end enterprise workflows with reduced human oversight, creating new digital workers across the enterprise. This means software evolves intelligently, making real-time adjustments to security, compliance, integrations, and business needs.

Balancing AppGen with agentic AI is key to ensuring that AI-generated applications become strategic assets that scale, integrate, and remain future-proof. As CIOs explore AI strategies, the question is how smartly AI-generated software can operate, evolve, and face the challenges of managing unstructured and structured data.

We shared Forrester’s perspective with our customers to see what IT leaders in various fields have to say about the future of enterprise IT, low-code, AppGen, and agentic AI.

In insurance, low-code is an opportunity for AI-enabled workflow orchestration

Frank Schmid, Chief Technology Officer (CTO) at General Reinsurance Corporation, sees opportunities rather than a dramatic end to applications as they exist today.

"I cast the challenge as an opportunity for low-code in orchestrating AI-enabled business workflow, rather than the death of the traditional app. The concept of dropping SaaS-based AI applications into a business workflow in the form of point solutions is not economically attractive. The adoption process of generative AI will play out over many years (if not, a few decades). So it will remain important to have an in-house AI engineering team that is capable of delivering integrated AI-enabled business workflow solutions using the PaaS model. This argument ties into the build vs. SaaS discussion raised by Bratincevic. The rise of AI agents in coordinating business workflows tilts the scales in favor of Platform as a Service as this enables the cloud engineering critical for orchestrating the AI tools of hyperscalers in business workflows."

Low-code is the first choice for universities

Howard Miller, CIO at the UCLA Anderson School of Management, explains why he thinks generative AI (gen AI) can enhance existing platform investments but not necessarily replace them in the near future.

"The build vs. buy discussion is interesting and complex, and I find that I agree with some points but not others. When I’m building new apps from scratch, I absolutely think low-code first. Gen AI is a game changer here. But I’m not sure it’s enough to swing the pendulum to one in which I build first and not buy. In fact, in some cases, key software on which I rely and on which I’ve already made a significant investment will absolutely benefit from having GenAI embedded. Even OutSystems falls into that category if we think about it. Generative AI should make what we’re using better. I’m just not sure at what price. Lastly, while I think the market is quickly changing, I’d suggest it’s more than five years for enterprises to entirely pivot from what they’re doing. Perhaps 10-15 is a semi-realistic target."

Consultants agree: AI just part of the solution

Andrew (Pete) Peterson, CTO at Riviera Partners, explains that human insights are critical to understanding what is needed from application development, no matter what organizations use to build them.

"When I was at the City of Oakland, we initiated the move away from large enterprise applications with our implementation of the OAKAPPS framework using low-code. The framework provided an infrastructure of common core functionality so that we could focus on building smaller functional applications specific to our business needs/requirements. This shift was more expedient, more efficient, and much more cost effective.Utilizing this framework, we were able to quickly deploy apps for Digital plan submission, Police Complaint Case Management, and Housing Assistance, to name a few. Enterprise applications are notoriously expensive, require costly consulting services and generally takes years to implement. Low-code platforms along with AppGen platforms significantly reduce the development time and cost of these smaller apps. AI also provides some ancillary business and functional knowledge, but the real business insight must come from internal business experts that intimately understand the processes the organization employs in these functional areas/scenarios."

Low-code remains critical in manufacturing sector

Cristiano Da Silva Marques, Director of Software Development at Vopak, believes that combining generative AI with the right team and other software development tools, generative AI can optimize application development.

"With a good multi-disciplinary team, companies are nowadays able to build fit-for-purpose solutions using low-code platforms. This is because the mix of technical and business knowledge is present and the right decisions are made in regards to the company's way of working. With generative AI, the decisions will become even easier to make, and when there is doubt in whether the products available are the right ones, prototyping ideas with AI can be almost immediate, which accelerates the innovation process. After that, low-code tools will allow you to evolve these applications to business-critical solutions with performance and architecture in mind."

AI, agentic AI, and application generation are part of the solution

No matter the industry, there is support for AI, agentic AI, and AppGen as part of a solution, rather than the entire solution. There are also arguments to be made for building and buying, but most agree that there is no one-size-fits-all approach. As firms consider whether to build customized solutions or buy ready-made ones tailored to their needs, the decision ultimately comes down to their specific bandwidth, budget, and use case.

Low-code platforms such as OutSystems allow firms to craft digital experiences that exceed expectations while eliminating the complexity of building custom software without the limitations of traditional code or COTS solutions. While the build vs. buy debate rages on, one thing is certain: We don’t need to write a eulogy for SaaS just yet.

*AppGen Is An Existential Threat To The Enterprise App Business
John Bratincevic, Principal Analyst
Nov 13 2024