Singapore – Dell Technologies has expanded its Dell AI Data Platform with a unified semantic layer, an enterprise knowledge graph and topic-specific agents, aiming to give AI systems consistent context drawn from a company’s own data.
The company also said new testing showed its data processing engine, running on NVIDIA GPUs, processes data nearly four times faster on average than CPUs alone.
Dell describes the platform as the data foundation of its Dell AI Factory. The additions target what it identifies as the main barrier to enterprise AI at scale: the data, rather than the model.
According to Dell, a company’s most useful information sits in files, databases, cloud services and systems across multiple sites. Most of it predates AI agents and was never labelled, connected or organised for a machine reader.
As a result, agents spend tokens reconstructing answers that should already exist, Dell says. Each query then costs more in steps and compute, and returns slower, less trustworthy results.
Arthur Lewis, President, Infrastructure Solutions Group, Dell Technologies, said, “Data without context is just noise. Most enterprises have spent years making their data accessible. That’s not the same as making it usable.”
“An agent that can find a customer record but doesn’t know what it means, how it connects to everything else, or whether it can be trusted isn’t intelligent. It’s just fast. The companies that solve it won’t just deploy AI faster or run more agents. They’ll get more out of the data they already have,” Lewis added.
The platform spans storage, orchestration, analytics and search, with each layer running on the same system. Dell says this means everything the platform produces traces back to governed data the customer already owns.
NVIDIA Nemotron Retriever models handle document parsing, embedding and reranking within the platform. NVIDIA cuVS, meanwhile, accelerates vector indexing and search.
Dell is adding three features to address a common weakness in current systems, which it says rebuild their understanding of data from scratch on every request. That forces them to work out what a term means and how it relates to everything else each time.
The first, the Unified Semantic Layer, gives structured and unstructured information consistent business meaning through rules, definitions and a searchable glossary. If one system refers to a client and another to an account, the AI can recognise them as the same thing.
Users can also import existing ontologies and classification taxonomies, whether industry-standard or enterprise-specific. Dell is additionally enabling NVIDIA Auto-Ontology, an open-source library that builds knowledge graphs from enterprise data, to extend the layer.
The second, the Enterprise Knowledge Graph, maps how structured and unstructured data relate so that an application or agent can locate the right context. It draws on metadata, lineage and query history to keep tuning the graph as activity changes.
When an agent asks a question, the platform pulls in every related item it is permitted to see, including tables, data products, multimodal data and vector indexes, wherever they sit. Dell offers the example of a manufacturer tracing an odd sensor reading to the machine, its repair history, the supplier batch and the orders at risk.
The third, Knowledge Agents, applies that context to specific tasks. Each acts as an adviser on a single topic, grounded in a defined slice of the knowledge graph.
Customers set the rules for each agent, including the guidance it follows, the data it can see, the quality bar it must meet and its spending limit. NVIDIA Nemotron Retriever models supply reasoning and visual understanding to the agents.
Dell says all three components hold sensitive information, so they remain inside the platform in the customer’s data centre. It adds that the platform does not lock customers into any single model, data or storage provider.
Beyond the context features, Dell has added performance and security upgrades to the data path, working with NVIDIA. The Dell Data Processing Engine, powered by NVIDIA cuDF on NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs, ran nearly four times faster on average than CPUs alone across a range of workloads, according to Dell’s testing.
The same testing recorded speeds of up to 20 times faster on batch processing workloads. Apache Arrow moves data between Dell storage and processing, which allows jobs to query data in place and can reduce data-preparation time, Dell says.
On storage, Dell PowerScale now supports up to 500 tenants in a single cluster. It also adds mTLS over NFS to encrypt and authenticate file traffic, alongside more granular role-based access control for each tenant.
Dell says this lets AI service providers and enterprises running shared AI platforms serve more teams and customers from one cluster while keeping each tenant’s data isolated.
Dell has also introduced the Dell Storage Performance Tool, which tests S3-compatible object storage across training, inference and checkpointing. The company says it helps customers size AI infrastructure and compare vendor solutions.
Separately, Dell is expanding implementation services for the platform across its data and storage engines. The services cover activating the analytics, processing, search and orchestration capabilities and keeping the platform tuned to AI workloads, according to the company.
Jason Hardy, Vice President of Storage Technology of NVIDIA, “AI agents are only as effective as the data they can access, understand, and trust, and NVIDIA accelerated computing and AI software help turn governed data into AI-ready context.”
“By bringing NVIDIA cuDF acceleration directly into the Dell Data Processing Engine, Dell helps shorten the path from stored data to GPU-accelerated AI-ready data their agents can use to drive real world innovation,” Hardy concluded.
