Dell expands its AI Data Platform so agents stop wasting tokens figuring out what your data means

Dell is expanding its AI Data Platform on the argument that, for most enterprises, the model isn’t the bottleneck; the data is.

Company information sits in files, databases, cloud services, and systems across multiple sites, and most of it was never labeled or connected with an AI reader in mind. The result is agents burning tokens trying to reconstruct answers that should already exist, which means more steps, more compute, higher cost, and less trustworthy results.

Dell’s answer is three new features designed to give data meaning and context, all of which stay inside the platform in the customer’s own data center and don’t lock users into a single model, data, or storage provider.

• Unified Semantic Layer: Gives structured and unstructured data consistent business meaning through rules, definitions, and a searchable glossary, so a term means the same thing everywhere. If one system says “client” and another says “account,” the AI knows they’re the same. Existing ontologies and taxonomies can be imported, and Dell is enabling NVIDIA Auto-Ontology, an open-source library that builds knowledge graphs from enterprise data.
• Enterprise Knowledge Graph: Maps how structured and unstructured data relates, using metadata, lineage, and query history to keep tuning itself. When an agent asks a question, it pulls in every related piece it’s allowed to see. Dell’s example is a manufacturer tracing one odd sensor reading to the machine, its repair history, the supplier batch, and the orders at risk.
• Knowledge Agents: Each acts as a trusted advisor on one topic, grounded in a defined slice of the knowledge graph. Customers set the guidance it follows, the data it can see, the quality bar it must clear, and how much it can spend. NVIDIA Nemotron Retriever models provide the reasoning and visual understanding.

On the infrastructure side, Dell is working with NVIDIA to build acceleration into the data path itself. Testing shows the Dell Data Processing Engine, powered by NVIDIA cuDF on RTX PRO 4500 Blackwell Server Edition GPUs, processes data nearly 4 times faster on average than CPUs alone, and up to 20 times faster on batch workloads. Apache Arrow moves data between Dell storage and processing so jobs can query data in place, which can cut data-preparation time.

Security and scale get upgrades as well. Dell PowerScale now supports up to 500 tenants in a single cluster, with mTLS over NFS to encrypt and authenticate file traffic and more granular role-based access control, so AI service providers and shared platforms can serve more teams from one cluster while keeping each tenant’s data isolated. The new Dell Storage Performance Tool lets customers test S3-compatible object storage across training, inference, and checkpointing, which helps with sizing AI infrastructure and keeping GPUs fed.

Availability

The post Dell expands its AI Data Platform so agents stop wasting tokens figuring out what your data means appeared first on Nasi Lemak Tech.