Joan Cros/NurPhoto via Getty Images
IBM is holding its annual THINK conference this week, and, unsurprisingly, artificial intelligence is the star of the show. The common theme across IBM’s broad swath of product unveilings is a focus on solutions that make it easier to scale enterprise AI in organizations, tackling challenges organizations face with AI deployment and integration.
AI agents
AI agents are the latest breakthrough in the AI space, taking the assistance that AI chatbots provide a step forward by actually performing tasks for people. Although agentic AI is a technology most enterprises should take advantage of, businesses face several implementation challenges, including finding ways to integrate it seamlessly into their various apps, data, and environments.
Also: Why scaling agentic AI is a marathon, not a sprint
To address these challenges, IBM unveiled a suite of enterprise-ready agents in watsonx Orchestrate. According to IBM, these AI tools enable businesses to build their own agents in under five minutes with both no-code and pro-code options; leverage pre-built agents for specialized use in specific domains such as HR, sales, and procurement; integrate with 80+ enterprise applications from the likes of Adobe, AWS, Microsoft, and more; orchestrate multi-agent, multi-tool coordination, and monitor agents with insights into performance, guardrails, governance, and more.
IBM also announced a new Agent Catalog in watsonX Orchestrate, which lets businesses more easily identify and access the best agent for their business use case from 150+ agents and prebuilt tools made available through partners and IBM’s offerings.
AI integration made easier
IBM also introduced webMethods Hybrid Integration, a solution designed to help enterprises integrate AI into their business operations with agent-driven automation. According to IBM, this makes it easier to manage “integrations across apps, APIs, B2B partners, events, gateways, and file transfers in hybrid cloud environments.”
Based on interviews with several companies using webMethods, an independent Forrester Consulting Total Economic Impact (TEI) study created a model of a typical organization representative of those customers. The study found that over three years, this composite organization experienced a 176% ROI, a 40% reduction in downtime, 33% time savings on complex projects, and 67% time savings on simple projects.
Tackling the data problem
Generative AI applications require a lot of data, and the efficiency of the AI model depends on the quality of that data. However, getting data into the ideal condition is often a challenge for businesses, as this typically takes lots of manual work to locate the unstructured data and then organize and structure it in a way that is most helpful for models.
IBM’s new watsonx.data seeks to help with that issue by combining an open data lakehouse with data fabric capabilities to help businesses unify and activate data across different formats and silos. According to IBM, wastsonx.data will help users connect their unstructured data with AI apps and agents, which would lead to 40% more accurate AI than when using the conventional RAG method.
Also: RAG can make AI models riskier and less reliable, new research shows
To further help enterprises work with unstructured data, IBM also launched watsonx.data integration, a single interface where users can access, manage, and work with data from different sources or locations, and watsonx.data intelligence, which leverages AI to yield deep insights from the unstructured data.
IBM also unveiled a new content-aware storage (CAS) capability available as a service in IBM Fusion with IBM Storage Scale. This capability can continuously analyze unstructured data and extract relevant information, which is then made available to RAG applications for faster processing.
Want more stories about AI? Sign up for Innovation, our weekly newsletter.
