At its annual Think conference, IBM announced significant updates to its Watsonx Orchestrate (WXO) platform, aimed at empowering developers to build and manage AI-powered agents for task automation. The new capabilities are designed to accelerate the development of agents that can take on repetitive tasks in the enterprise, marking a shift from AI assistants to AI agents.

According to IBM CEO Arvind Krishna, the updates are part of the company's strategy to capture a larger share of the growing market for generative AI applications. As enterprises are expected to double their AI investments over the next few years, IBM is positioning WXO as a key platform for building and deploying AI agents that can complete tasks without manual intervention while learning from new data.

The updated WXO platform features a new no-code studio called Agent Builder, which enables developers to create agents using drag-and-drop tools. Additionally, developers can import outside tools and automations, such as open-source frameworks, third-party or custom tools, to build custom agents. While IBM is not the first to offer no-code tools for agent creation, its Agent Builder is designed to provide a simple and unified user experience optimized for scalability.

Other additions to WXO include a new agent development kit (ADK) for professional developers, which allows them to create highly specialized agents from scratch. IBM is also introducing an Agent Catalog, featuring over 150 pre-built agents developed by IBM and its partners, including Box, MasterCard, Oracle, Salesforce, and ServiceNow. These agents can be embedded in systems for web research, performing calculations, or other tasks, and can also be used as templates for creating new agents in Agent Builder.

IBM plans to offer revenue-sharing agreements to supporting partners for agents they build with WXO and publish in the catalog, although details of the agreements were not disclosed. The company is also playing catch-up in this area, as vendors like Salesforce, ServiceNow, and Microsoft already offer similar marketplaces with their agent-building platforms.

Another significant update to WXO is the introduction of a multi-agent orchestrator, which enables the orchestration of multiple agents, whether built by IBM, a customer, or a third party. This capability allows agents to share information and tackle complex, multi-step processes together, making it easier to automate workflows and tasks.

WXO also includes a set of observability tools to help developers monitor their agents, including capabilities for monitoring AI performance, reliability, and enforcing AI guardrails. The platform is generally available, with pricing based on consumption, and an "Essentials" version offering entry-level pricing for clients just starting out with WXO.

According to Gartner vice president analyst Jim Hare, while IBM's updates are significant, it's unlikely that WXO will gain traction outside of customers already part of the IBM ecosystem and using other IBM products, including other Watsonx products. However, there may be exceptions, such as customers requiring hybrid, on-prem, or edge deployments.

As the market for AI-powered task automation continues to grow, IBM's updates to Watsonx Orchestrate position the company as a key player in the space. With its focus on ease of use, scalability, and flexibility, WXO is well-positioned to help enterprises automate tasks and workflows with AI agents, driving greater efficiency and productivity.