Agents are rapidly advancing in automating digital work, but enterprises face a harder challenge: moving beyond prototypes to deployed systems that deliver measurable business value. This path is complicated by fragmented frameworks, slow development, and the absence of standardized evaluation practices. Generalist agents have emerged as a promising direction, excelling on academic benchmarks and offering flexibility across tasks, applications, and modalities. Yet, evidence of their use in enterprise settings remains limited. This paper reports IBM’s experience developing and piloting the Computer Using Generalist Agent (CUGA). CUGA adopts a hierarchical planner--executor architecture with strong analytical foundations, achieving state-of-the-art performance on AppWorld and WebArena. Beyond benchmarks, it was evaluated in a Business-Process-Outsourcing talent acquisition pilot, addressing enterprise requirements for scalability, auditability, safety, and governance. In preliminary evaluations, CUGA approached the accuracy of specialized agents while suggesting reductions in development time and cost. We provide early evidence that generalist agents can operate at enterprise scale, distill key technical and organizational lessons, and outline requirements for transitioning research-grade architectures like CUGA into enterprise-ready systems.

We propose to develop a framework which provides the ability to apply complex event processing in realtime domains, thus allowing an easier process of developing and maintaining specific solutions for real-time event-based systems, while upholding the real time requirements of the system. Specifically, we propose to develop a framework that includes an integrated development environment for defining rules, and, given a set of rules, generates code for a complex event processing application for which it is able to determine time bounds on the response of this application to a set of supported events. In particular, the tool helps determine a time bound for the execution time of the code corresponding to each rule. Many Service Oriented Architecture (SOA) applications, in domains such as financial services, manufacturing, gaming and military/aerospace, have real-time performance requirements. We present real-life industry use cases from these domains as motivation for the potential benefit in developing real-time complex event processing applications. In support of a feasibility argument for the proposed approach we present some preliminary experimental results obtained on a partially implemented tool.

Events are at the core of reactive and proactive applications, which have become popular in many domains.