OutSystems has launched an open Agentic Systems Platform for enterprise AI designed to reduce customer dependence on any single large language model provider.
The launch also includes a new Agent Experience layer, a broader partnership with AWS, and a distributed architecture that lets customers run AI workloads in different environments. The changes target organisations that want tighter control over how AI systems are built, governed, and deployed.
OutSystems presented the platform as a way for businesses to separate their business logic and data from the underlying AI models and tools they use. The approach reflects wider concern among large companies over vendor concentration, regulatory demands, and the cost of switching AI suppliers as the market develops.
The platform is built around what OutSystems calls an Enterprise Context Graph, which supports orchestration and governance across AI agents, applications, and operational data. It also introduced the OutSystems Agent Experience, which exposes a set of A2A and MCP tools and services for developers building and managing agent-based systems.
The first services released with that layer cover agentic coding, publishing, and platform extensibility. They are intended to support developers creating and governing portfolios of AI agents inside large organisations.
Partnership expansion
AWS features heavily in the latest product line-up. OutSystems has added a native integration with Kiro, an agentic development environment from AWS, to its engineering tools for building, managing, and governing agents and applications on the platform.
It also launched Legacy Modernization Services using AWS Transform, creating an automated pipeline for migrating older systems based on technologies including COBOL and Lotus Notes into newer agent-based systems. Those services are in preview for selected customers.
For enterprise orchestration, the latest release is backed by Amazon Bedrock. OutSystems described this as the next stage of its Agent Workbench product, adding agent evaluations, guardrails, semantic search, and broader Bedrock support for customers already using the software.
The partnership also has a geographic and infrastructure element. OutSystems said the expanded work with AWS increases the platform's reach, while additional European sovereign cloud providers are expected to be added. The company also introduced full runtime isolation and self-hosting options, which it said would help customers place AI workloads where sovereignty requirements demand.
Customer examples
OutSystems used customer examples including the Dutch Red Cross, Butler Plus, and AllianceCorp Manufacturing to show how clients are adopting agent-based processes. It presented those deployments as evidence that customers are moving beyond narrow AI use cases toward broader operational systems.
Woodson Martin, Chief Executive Officer at OutSystems, said the company sees the market shifting toward greater demand for choice and control.
"Enterprise leaders are realizing they need leverage and control in this highly fragmented and dynamic AI market. To maintain that leverage and protect margins, organizations must separate their proprietary business logic and data from specific AI providers," said Woodson Martin, Chief Executive Officer, OutSystems.
"Our Agentic Systems Platform provides the open and neutral platform to ensure optionality and control. Customers can optimize for digital sovereignty, performance and cost and can work across all the latest models and tools without rebuilding their core operations every time. In a time of great change, we are ensuring the CEO and CFO maintain control over the company's future," Martin said.
One customer also pointed to the use of governed AI agents in core processes.
"The ability to orchestrate an agentic enterprise will transform our operations. We are no longer just using AI for isolated tasks; we are deploying governed agents that handle mission critical processes, while maintaining strict compliance," said Mike Schmitt, Chief Executive Officer, Butler Plus.
Industry focus
Beyond platform and tooling updates, OutSystems introduced a banking product for loan origination, the first in a planned set of industry-specific offerings. The company said the product is intended to make it easier for financial institutions to deploy AI systems tied to specific workflows.
That banking offering also uses Amazon Bedrock, according to OutSystems, and is linked to domain-specific models for sectors such as finance and law, as well as narrower tasks including document processing. A related banking and financial services agent kit is already available for developers.
The company, founded in 2001, said its ecosystem includes more than 85 million end users, more than 600 partners, and customers in more than 75 countries. The latest launch suggests OutSystems is positioning itself less as a conventional app development software supplier and more as a control layer for enterprises seeking to manage AI agents, legacy modernisation, and sector-specific deployments on the same platform.