Rapid agentic AI at scale possible with new Teradata solution

Teradata has announced the introduction of Enterprise Vector Store to enable organisations to deploy Trusted agentic AI at scale in a more cost-effective manner and with increased flexibility.

Enterprise Vector Store has been designed to offer cost-effective and rapid response times across various data volumes, addressing complex problems through the integration of structured and unstructured data. In future developments, it is planned to incorporate NVIDIA NeMo Retriever microservices, a component of the NVIDIA AI Enterprise software platform.

According to Teradata, this new offering allows processing of billions of vectors with integration into existing enterprise systems, achieving response times as fast as tens of milliseconds. It aims to provide businesses with the necessary sophistication to address complex challenges cost-effectively.

The Enterprise Vector Store creates a unified and reliable repository for all data and extends the support Teradata currently offers for retrieval-augmented generation (RAG). This forms part of its broader goals to develop dynamic agentic AI use cases, such as the "augmented call center."

CTO of Teradata, Louis Landry, commented, "Vector stores are at the root of how we bind truth to generative AI models and agentic AI. They are essential to any data management practice, but their impact is limited when they are slow or siloed. Teradata's long-standing expertise in high concurrency and linear scale, as well as the critical ability to harmonize data and support RAG, means Teradata Enterprise Vector Store delivers on the dynamic, trusted foundation large organizations need for agentic AI."

This product is intended to support use cases that necessitate vector capabilities and RAG applications effectively. It promises cost-efficient scaling and easy integration, offering enterprises a way to derive value from unstructured data while reducing expenses.

Enterprise Vector Store manages unstructured data across various formats — text, video, images, PDFs, and more — ensuring a cohesive analysis by unifying structured and unstructured information. It involves the entire lifecycle of vector data management, including embedding generation, indexing, metadata management, and intelligent search.

Integration within the existing Teradata system offers versatile deployment options, including cloud, on-premises, or hybrid models. The solution also plans to add temporal vector embedding, aiming to enhance trust and explainability by monitoring changes to data over time.

Pat Lee, Vice President of Strategic Enterprise Partnerships at NVIDIA, stated, "Data is essential to accurate inference for AI applications. Teradata Enterprise Vector Store, integrated with NVIDIA AI Enterprise and NVIDIA NeMo Retriever, can unlock the institutional knowledge stored in PDFs and other unstructured documents to power intelligent AI agents."

The augmented call center is a highlighted use case for the Enterprise Vector Store, demonstrating the application of agentic AI and RAG in transforming customer service. This involves using AI agents to deliver personalised responses and upsell or cross-sell opportunities during customer interactions.

Through examples such as an insurance company, the solution illustrates how vector data management can provide context-aware responses to customer inquiries promptly. This is achieved by leveraging the harmonised data and NVIDIA NeMo Retriever's capabilities to analyze contracts and offer pertinent recommendations.