IBM is introducing new hybrid technologies to help companies scale up AI. The solutions help build and implement AI agents using business data. IBM’s Watsonx Orchestrate makes it possible to build AI agents that work with more than 80 business applications in just five minutes.
During its annual THINK event, IBM announced several innovations to help companies scale their AI initiatives. The tech giant estimates that by 2028, there will be more than a billion apps, forcing companies to integrate seamlessly into increasingly fragmented IT environments.
Build AI agents in five minutes
To support this, IBM is introducing enterprise-ready agent capabilities in Watsonx Orchestrate. These capabilities allow companies to build their own agents in less than five minutes with tools that simplify integration, customization, and deployment—from no-code to pro-code solutions for different users.
In addition, IBM offers pre-built domain agents for areas such as HR, sales, and procurement and simpler utility agents for basic tasks such as web research. Integration with more than 80 business applications, including Adobe, Microsoft, Oracle, Salesforce, SAP, and ServiceNow is a key advancement.
Integration remains a major challenge in the adoption of AI. Most organizations rely on a patchwork of APIs, apps, and systems spread across on-premise and multi-cloud environments. IBM addresses this with the new webMethods Hybrid Integration. This solution replaces rigid workflows with intelligent, agent-driven automation.
Leveraging unstructured data
Unstructured data—hidden in contracts, spreadsheets, and presentations—is a valuable but underutilized company resource. IBM has further developed watsonx.data to help organizations activate this data for better AI applications. The updated watsonx.data brings together an open data lakehouse with data fabric capabilities to unify data across silos, formats, and clouds.
Tests show that this approach can lead to 40 percent more accurate AI than conventional RAG (Retrieval-Augmented Generation). In addition, IBM is introducing watsonx.data integration for orchestrating data and watsonx.data intelligence for extracting insights from unstructured data.
IBM is also launching IBM LinuxONE 5, a Linux platform for data, applications, and reliable AI that can process up to 450 billion AI inference operations per day. The new LinuxONE innovations include advanced AI accelerators with IBM’s Telum II on-chip AI processor and the IBM Spyre Accelerator.