Box to embed Nvidia Nemotron in content retrieval
Original title: Box taps Nvidia’s Nemotron for new agent-first content retrieval system
Enterprise Advance customers get it early next year; Nvidia expects faster queries and 90% fewer tokens.
At its developer conference, Box said Box Content Retrieval will use Nvidia’s Nemotron embeddings so agents get relevant document chunks instead of reading entire files. CEO Aaron Levie said the API combines Box metadata and permission controls with several retrieval technologies. Nvidia’s Kari Briski told TechCrunch she expects queries four times faster, with 20% fewer errors and 90% fewer tokens. The feature is for Enterprise Advance customers and is due early next year.
Unconfirmed: The speed, error, and token figures are expectations from an Nvidia executive, not published measurements.
- Nemotron Embed offers open-weight models for automated retrieval.
- Box says Nvidia’s July Nemotron release topped the RTEB benchmark.
- Box cites lawyers finding clauses and precedents as a use case.
- Levie said the API packs about five technologies together.
- Products
- Box Content Retrieval
- AI models
- Nemotron
From the original article
2 min read at techcrunch.com
The system can answer queries four times faster, with 20% fewer errors and 90% fewer tokens used.
By Russell Brandom · techcrunch.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.