Powered by OutSystems
After two years of flashy AI demos, rushed agent prototypes, and breathtaking predictions, enterprise technology leaders are striking a more realistic tone for 2026. In a recent webinar hosted by OutSystems, a panel of software executives and enterprise practitioners argued that the most important AI efforts taking place today are focused on the practical issues of governance, orchestration, and iteration, and on integrating agents into the systems that have taken decades to build.
Corporate leaders are increasingly focused on fundamentals. Utilization of new AI technology is a priority
To accelerate productivity, improve delivery times, and generate measurable business results.
Three elements shape this work:
-
Moving from an AI agent prototype to an agent system that delivers measurable ROI in production
-
Enterprise platforms will play a growing role in securely managing, orchestrating, and scaling AI agents
-
The rise of generalist developers and enterprise architects as the most valuable technical profiles in the era of AI-generated code
Against this backdrop, the panel discussed governance frameworks, the economics of enterprise AI, and the limitations of large-scale language models without orchestration. The conversation eventually turned to how leading organizations are building multi-agent systems based on existing enterprise data and workflows.
real world agent
To enable agents to work in production across your enterprise, it’s best to use an integrated platform that handles development, iteration, and deployment. That’s where features like the OutSystems platform’s agent workbench become important, said Rajkiran Vajreshwari, senior manager of app development at Thermo Fisher Scientific. It provides the infrastructure to learn, iterate, and manage agents at scale.
His team at Thermo Fisher transitioned from single-task AI assistants in customer service to using Workbench to build a coordinated team of expert agents. When a support case arrives, the triage assistant categorizes the request and dynamically routes it to the appropriate specialized agent, such as an intent and priority agent, product context agent, troubleshooting agent, or compliance agent.
“You don’t have to think about how anything works. It’s all pre-built,” he explained. “Each agent has a narrow role and clear guardrails. They are accurate and auditable.”
Manage shadow AI risks
AI creates new categories of risk when anyone in a company can generate production-grade code without IT oversight. Essentially, this is an ungoverned shadow AI. These homegrown products are prone to hallucinations, data leaks, policy violations, model drift, and agents performing actions that are not officially authorized.
Luis Blando, CPTO at OutSystems, says leading organizations need to do three things to get ahead of this risk.
“Give users guardrails. Users are going to use AI whether they like it or not. Companies that appear to be making progress are using AI to manage AI across their portfolio,” he explained. “That’s the difference between shadow AI disruption and enterprise-level scale.”
Eric Kavanagh, CEO of The Bloor Group, said governance requires a hierarchical set of disciplines, including protecting data, monitoring model drift, and making intentional choices about where AI connects with existing business processes.
“Companies don’t have to manually create these controls,” he added. “Many of these guardrails and levers are built into platforms like OutSystems.”
Why the real challenge for orchestration is the model and platform
Much of the early excitement around enterprise AI focused on choosing the right large-scale language model. Today, the more difficult challenge, and the far more enduring source of value, is orchestration. This includes routing tasks, coordinating workflows, managing execution, and integrating AI into existing enterprise systems.
Scott Finkle, vice president of development at McConkey Auction Group, said that no matter how good an LLM is, it is part of a complex workflow, not the final solution. Organizations should be prepared to hot-swap between Gemini, ChatGPT, Claude, and whatever comes next without having to rebuild their agent systems.
Platforms with orchestration capabilities make this possible. Even when AI handles the top inference layer, it manages the lifecycle, provides visibility, and ensures processes are executed reliably.
“AI and models may change and workflows may change, but the orchestration remains the same,” Finkle says. “That’s how we derive value from AI.”
The economics of enterprise AI investing
Security, compliance, governance, and platform-level AI capabilities will all require greater investment in 2026, especially as AI moves into core workflows such as finance and supply chain. Companies should prioritize incremental gains rather than expecting big profits right away.
“We’re focusing on the bases,” Finkle said. “The key is to get something into production and make an impact. You’re not going to save money by investing a lot of money in a pilot project that won’t go into production. It won’t happen overnight, but over time I think the savings will be huge.”
Opinions remain divided on how companies should approach AI transformation. Some people start from scratch and rethink the entire process. Companies also want to integrate AI into their systems, especially those with billions of dollars of existing infrastructure depreciating internally. They want to enable agent systems to reuse data, APIs, and proven processes while accelerating delivery. The agent platform approach helps both camps, but especially the latter. Organizations can deploy agents that add tangible value while maintaining the integrity of established, deterministic workflows.
The rise of enterprise architects and generalist developers
As AI accelerates code generation, software delivery bottlenecks are disappearing. Instead, the emphasis is on systems thinking. This is the ability to understand the broader enterprise architecture, decompose complex business problems, and reason about how AI integrates with existing infrastructure. Kavanagh specifically singled out enterprise architects as the professionals best positioned to make the most of this moment.
“We are entering a very interesting time for generalists,” he explained. “The more you know about your enterprise and business architectures and how they work together, the better off you will be.”
“The result is faster delivery with fewer interruptions and bugs,” Kavanaugh said. “You can focus on non-repetitive tasks, which is a win-win for developers, the business, and the entire IT organization.”
Watch the entire webinar here.
Sponsored articles are content created by companies that pay us to post or have a business relationship with VentureBeat, and are always clearly marked. For more information please contact us sales@venturebeat.com.
