IBM this month unveiled software that radically simplifies the enterprise data stack, integrating, managing and leveraging the unstructured enterprise data needed to power AI agents and other advanced AI applications.
The new products include IBM watsonx.data integration and IBM watsonx.data intelligence , some of whose capabilities are also available through watsonx.data , IBM’s open, hybrid data lakehouse , designed to help companies manage the entire lifecycle of data for AI in one place.
These new software are hybrid, open and can be integrated with third-party data stacks, providing flexibility and interoperability to drive innovation across the ecosystem. Internal validation with IBM watsonx.data has shown the potential to deliver AI that is 40% more accurate than traditional RAG (Search Augmentation Generation).
Also Read: UiPath debuts Enterprise-Grade Agent automation platform
IBM client Lockheed Martin is using the new watsonx.data to empower its 70,000 engineers, scientists and technicians to get answers and information from millions of documents in natural language. “By rapidly deploying solutions from the lab to the field, we’re accelerating innovation and efficiency to help make the world safer and more secure,” said John Clark, Lockheed Martin’s senior vice president of technology and strategic innovation.
Businesses need to adopt generative and agent-based AI to drive innovation, improve productivity and stay competitive. And to use AI with precision, they need company-specific data. According to a recent CEO survey by IBM, 72% of CEOs believe their company’s unique data is the key to unlocking the value of generative AI.
However, much of this important data is unstructured and buried in emails, PDFs, presentation materials, videos, etc., making it difficult to utilize. Conventional RAG cannot handle the scale and complexity of unstructured data, and it is not easy to properly integrate it with structured data. In addition, there are a wide variety of tools for data management, and the complexity of the data stack is an issue.
As a result, unstructured data, which according to IDC represents up to 90% of an enterprise’s data, is currently underutilized and not reflected in AI agents and other generative AI applications.
SOURCE: PRTimes
