6 May 2025
Emerging AI agents represent a fundamental tipping point in the AI revolution, shifting from AI that can chat with you, to AI that can do work for you—with unprecedented autonomy.
For decades, interacting with powerful enterprise systems required specialized knowledge to navigate a vast network of complex interfaces. AI agents will lower the barrier to accessing AI's power. By enabling interaction through conversational interfaces, users can simply state their goals—while networks of agents take on the necessary actions across backend systems. This democratization means every person, regardless of their technical expertise, can leverage AI to automate complex tasks, gain insights and drive productivity across functions like HR, finance, IT and customer service.
We believe this new operating model will be the catalyst that ignites the massive economic potential anticipated for generative AI—which has been estimated to generate up to USD 4.4 trillion economic value annually across industries, according to McKinsey.
But to make it a reality, agents must be able to work seamlessly across the vast web of applications, data and systems that underpin today's complex enterprise technology stacks. Which means that orchestration, integration and automation are the secret weapons that will move agents from novelty into operation.
That’s why IBM isn’t just making agents—we're re-imagining the underlying technology stack that will power the agent era.
IBM’s strategy is all about helping clients put agents to work across all their technologies, on any infrastructure, powered by all their data. Our portfolio enables businesses to orchestrate agents from any vendor and platforms like Salesforce Agentforce, while also leveraging the power of their existing IT estate–including technologies from Adobe, Microsoft, Oracle, ServiceNow, Workday and more. And to do it quickly, with pre-built agents, tools and integrations.
Today, we’re unveiling the most comprehensive set of agent capabilities in the industry to help companies build, run and manage agents—delivered via IBM watsonx Orchestrate.
Core to this strategy is IBM watsonx Orchestrate, our unified solution to orchestrate work across your entire enterprise landscape. It functions as a multi-agent supervisor, router, and planner, enabling agents–whether from IBM, third-party platforms or open source–to collaborate effectively with each other and with your existing assets (traditional automations, AI assistants, APIs, data stores, and applications). It’s pre-integrated with 80+ leading enterprise applications, making it the ideal gateway for infusing agents across the full enterprise stack—no vendor lock-in.
IBM will offer pre-built agents with specific expertise and skills in business functions that are prime targets for productivity gains. Our first domain agents to be introduced are:
We plan to roll out additional domain agents in areas like customer care and finance throughout the year. Beyond specialized domain agents, we’ve also developed a wide range of general utility agents that can be embedded within systems for tasks like web research or calculations.
To simplify access to agents from IBM and other vendors, we’re excited to introduce our upcoming Agent Catalog for watsonx Orchestrate.3 This will serve as the launchpad for accessing hundreds of pre-built tools and agents from IBM and our vast ecosystem of partners, with more than 150+ already included. Some early examples4 include a sales prospecting agent with Salesforce that enables sellers with the tools and knowledge to effectively prospect and increase sales conversion rates, and an HR agent that can be embedded in Slack to handle employee queries via Slack’s conversational interface.
We’re also launching a new program which will allow any IBM partner to build directly with watsonx Orchestrate, helping expand our portfolio of partner-built apps with even more speed and ease.
IBM is also focused on helping companies build, integrate and customize their own custom agents for a wide range of skillsets, using any framework or architecture - including open-source agent frameworks like CrewAI, LangGraph, and IBM Bee.
Our new watsonx Orchestrate Agent Builder5 available in public preview provides business users with a no-code automation studio that allows them to build their own agent in 5 minutes, tapping into pre-built agents and tools from our Agent Catalog. This studio democratizes the process of agent building, empowering anyone in the organization to easily build, deploy, and manage sophisticated agents grounded in business knowledge, leveraging hundreds of pre-built tools.
For developers who need control and flexibility, we offer a wide set of pro-code toolsets and software development kits to build, evaluate and deploy custom AI agents. This includes a new agent development kit (ADK) to build highly specialized agents from scratch and deploy them directly from your laptop into watsonx Orchestrate. In addition to our suite of pre-built tools and integrations with core business applications, we are also expanding agent connectivity with full support for the Model Context Protocol (MCP), allowing AI agents to integrate easily with thousands of existing tools.
The true power of AI agents in the enterprise is unlocked when they work in concert. As agents evolve to take on even more complex enterprise projects, orchestration capabilities will be essential for enabling agents to work together within and across essential business workflows.
Recognizing this, we are enhancing watsonx Orchestrate with multi-agent orchestration capabilities—moving beyond simple task routing to enable advanced, agent-to-agent collaboration. This allows specialized agents, whether built in-house, by partners, or using open source, to share information, and tackle complex, multi-step processes together. Companies can embed these capabilities into their agentic systems to analyze user requests and route instructions across the right agents, assistants, and skills in real time.
But effective orchestration isn’t just about orchestrating agents–it's also about leveraging your entire technology landscape. Which is why watsonx Orchestrate enables collaboration across agents and with your existing technology investments–existing automations, APIs, data sources, and core applications. This way, agents can operate as a value-add within your business operations, rather than operating in isolation.
As agents begin to proliferate, tools for agent observability will be essential for effective and responsible adoption of these increasingly autonomous systems. This includes the ability to discover, manage, monitor, and optimize the use of agents across the enterprise.
Our watsonx portfolio offers a robust set of tools to monitor AI performance and reliability, to enforce AI guardrails, and to help scale AI resources effectively. For instance, we have tools that help companies evaluate and select which AI models to use based on specific goals like cost-efficiency or performance. We’re also developing tools that help with the discovery and orchestration of agents and tools, enabling tasks to be designed and executed more effectively. Additionally, our industry-leading AI governance capabilities can give companies greater visibility and control over agentic systems, helping manage accuracy, performance and risk.
Learn more about IBM’s offerings to help companies build, deploy and manage AI agents and assistants.
Explore watsonx Orchestrate, and sign up for a free trial.
Disclaimers:
Statements regarding IBM's future direction and intent are subject to change or withdrawal without notice, and represent goals and objectives only.
Footnotes:
1 Available today in public preview, planned GA in June
2 Available today in public preview, planned GA in June
3 Available today in public preview, planned GA in June
4 These examples are expected to be available in June
5 Available today in public preview, generally available in June
What's New at IBM newsletter
Get the biggest product and feature announcements, including recent video chats on products, and educational offerings from IBM and our training partners. See the IBM Privacy Statement.