At its developer conference on Thursday, Box unveiled a new capability that integrates Nvidia's Nemotron embedding system directly into the heart of its platform. The feature, dubbed Box Content Retrieval, enables agents to respond to queries much more effectively by leveraging the embedded metadata generated by Nemotron. According to Box, users will benefit from faster, more accurate, and cheaper queries. The system will be offered to all of Box's Enterprise Advance customers, with rollout planned for early next year.

"This dramatically boosts both the relevance of our search technology for agents," said Box CEO Aaron Levie. "Instead of an agent having to kind of read through each document, we can send back chunks from the documents that are most relevant to the user's query or whatever the task is that the agent is doing."

The feature represents an implementation of Nvidia's Nemotron Embed project, which provides open-weight embedding models designed to enhance automated retrieval systems. Such models—which use LLMs to pre-process data so agents can use it more effectively—have often been neglected by major labs. This area has been a key focus for Nvidia's open-source model initiative, and its recent Nemotron release in July claimed the top spot on the RTEB benchmark, which measures retrieval quality.

Box notes that the system delivers the most value when paired with active agents, citing the example of lawyers who will ostensibly be able to locate relevant clauses and precedents more easily under the new system. Since Nemotron's embeddings move beyond simple keywords to map underlying concepts, a search tool can surface related legal themes even when different wording or language is used. While the Nemotron models are freely available, the system also depends heavily on Box's pre-existing metadata and permission controls—factors that would make it difficult to recreate from scratch.