OutSystems has launched Agent Workbench, a new platform aimed at simplifying the development and deployment of artificial intelligence agents for businesses. The platform, announced at the company’s annual conference in Lisbon, is designed to allow organizations to create and customize AI agents without the need for specialized engineering teams. This initiative addresses a growing demand in the technology sector, as many companies are exploring the potential of AI to automate tasks and enhance their operational efficiency. The general availability of Agent Workbench follows an early access program that included major companies such as Axos Bank and Thermo Fisher Scientific, which used the platform to build and test various AI-driven solutions.
The core of Agent Workbench is its low-code approach, which abstracts away much of the complexity typically associated with AI development. This enables businesses to build custom AI agents that can perform a wide range of functions, from analyzing log data to managing customer service inquiries. By providing a unified platform for data, workflows, and AI agents, OutSystems aims to help organizations overcome common hurdles in AI adoption, such as governance, security, and integration with existing systems. The platform supports a variety of large language models (LLMs) from providers like AWS Bedrock, Azure OpenAI, and Anthropic, as well as open-source models, allowing for flexible integration with a company’s existing technology stack.
Addressing Key Enterprise Challenges
According to OutSystems, a significant majority of organizations are prioritizing the development of AI agents but are encountering significant obstacles. Research cited by the company indicates that 93% of organizations are focused on AI agent development, yet they face challenges related to governance, security, integration, and scalability. OutSystems CEO Woodson Martin stated that many companies are stuck in endless pilot programs and are dealing with a proliferation of ungoverned tools, which stalls business impact. Agent Workbench is designed to address these issues by providing a centralized and secure environment for building, deploying, and managing AI agents.
The platform aims to move companies beyond experimental AI projects to creating true business enablers. Martin emphasized that legacy systems, siloed data, and fragmented AI tools create complex development cycles that slow down progress. By offering a unified platform, OutSystems intends to streamline the entire process, from development to deployment, and provide the necessary guardrails for governance and security. This approach is intended to allow businesses to realize the benefits of AI more quickly and efficiently, without needing to make substantial investments in specialized AI talent.
Early Adopter Success Stories
Several major companies participated in the early access program for Agent Workbench and have already implemented AI agents in their operations. Axos Bank, for example, used the platform to automate log analysis and document processing tasks. Kevin Hearn, SVP and Head of Consumer Bank Development at Axos Bank, said the company plans to expand its use of AI capabilities within OutSystems to achieve immediate gains without needing to hire specialized AI personnel. Axos is also developing agents to analyze error logs and automate data entry from documents.
Thermo Fisher Scientific, another early adopter, deployed a customer escalation agent to interpret unstructured data from customer interactions, thereby eliminating manual triaging processes. In the United Kingdom, Arch Company, the largest landlord for small businesses, implemented a content classification agent to route customer service inquiries automatically, without manual intervention. These examples illustrate the practical applications of Agent Workbench in various industries, from financial services to scientific technology and real estate. The feedback from these early adopters was instrumental in refining the platform before its general release.
Technical Capabilities and Integrations
Agent Workbench includes several features designed to facilitate the creation and management of AI agents. The platform offers an agent marketplace and supports the Model Context Protocol, which allows agents to connect with enterprise systems and external tools directly. This capability is crucial for building agents that can perform complex tasks that require access to multiple data sources and applications. The platform’s support for a wide range of LLMs is another key feature. It is compatible with models from major cloud providers such as AWS Bedrock and Azure OpenAI, as well as models from Anthropic, Google (Gemini), Cohere, Mistral, Databricks, AI2, and IBM’s watsonx.
Furthermore, Agent Workbench accommodates custom-built models on VertexAI and open-source models from HuggingFace, providing a high degree of flexibility for companies with specific needs or existing AI investments. This extensive model support simplifies the integration of Agent Workbench into a company’s existing IT landscape, as it can work with a variety of AI models and data sources. The platform’s low-code nature means that developers do not need deep expertise in AI to build and deploy agents, which can significantly accelerate the development process.
Expanding Enterprise Use Cases
Beyond the initial early adopters, other companies are also leveraging Agent Workbench to develop innovative AI solutions. TeamWork, an international consulting group, built a multi-agent system that provides real-time guidance to support teams while automating routine ticket resolution. The company anticipates that a significant portion of tickets will be resolved automatically, with a considerable decrease in resolution times for more complex cases. In India, Grihum Housing Finance is using the platform to improve the accuracy of its loan underwriting process by analyzing property evaluation reports and identifying technical property deviations.
The platform has also been tested by KPMG’s Low-Code Centre of Excellence. Hélio Pimenta, an Associate Partner at the center, noted that Agent Workbench provides the speed-to-value and governance guardrails that organizations expect from OutSystems development. These use cases demonstrate the versatility of the platform and its potential to be applied to a wide range of business processes, from customer service and IT support to financial analysis and risk management. The ability to customize agents to specific business requirements is a key benefit that enables companies to address their unique challenges and opportunities.
The Rise of Agentic AI
The launch of Agent Workbench comes at a time when agentic AI is emerging as a significant trend in the technology industry. AI agents are systems that can perform tasks autonomously, often by interacting with other systems and data sources. They have the potential to automate a wide range of business processes, from simple data entry to more complex decision-making. The ability to create and deploy these agents at scale could provide a significant competitive advantage to businesses, enabling them to improve efficiency, reduce costs, and deliver better customer experiences.
By providing a low-code platform for building AI agents, OutSystems is aiming to make this technology more accessible to a broader range of organizations. The company’s approach of combining a user-friendly development environment with robust governance and integration capabilities is designed to address the key challenges that have so far limited the adoption of agentic AI. As more businesses look to leverage the power of AI, platforms like Agent Workbench are likely to play an increasingly important role in enabling them to build and deploy intelligent, automated systems.
