Abstract

Enterprise Retrieval Augmented Generation (RAG) is only as effective as the enterprise context it can access. As organizations move from proof-of-concept to production, the ability to efficiently create, maintain, and refresh trusted, retrieval-ready context becomes critical to improving retrieval accuracy, accelerating indexing, and supporting growing data volumes and user demand.

Building on the Enterprise RAG deployment described in the Enterprise RAG Deployment Guide v2.0, this document explores how to scale the data extraction and ingestion pipeline used to transform enterprise data into high-quality context for AI applications and agents. Specifically, it provides guidance and insights for scaling Enterprise Retrieval Augmented Generation (RAG), based on NVIDIA’s RAG Blueprint, and the NeMo Retriever extraction application built using the Nemo Retriever Library (NV-Ingest) on the NVIDIA Enterprise Reference Architecture for NVIDIA RTX PRO™ 6000 Blackwell Server Edition and NVIDIA H200 NVL systems.

This is the second guide in a series of 3 documents covering the deployment, ingestion, and retrieval stages of Enterprise RAG:

• Enterprise RAG Deployment Guide v2.0

• Enterprise RAG Ingestion Scaling Guide v2.0

• Enterprise RAG Retrieval Scaling and Sizing Guide v2.0