The Red Hat Ecosystem Catalog is the official source for discovering and learning more about the Red Hat Ecosystem of both Red Hat and certified third-party products and services.
We’re the world’s leading provider of enterprise open source solutions—including Linux, cloud, container, and Kubernetes. We deliver hardened solutions that make it easier for enterprises to work across platforms and environments, from the core datacenter to the network edge.

Enterprise-grade RAG on complex unstructured docs. Built for air-gapped, on-premises, and managed cloud deployments.
The world's most important documents don't live on the public net and never will. GroundX Rag On-Prem is built for regulated industries who want to deploy retrieval augmented generation where their data lives… in air gapped or connected data centers or private cloud, with enterprise-grade accuracy, scalability and security.
GroundX is feature complete RAG: ingest, parse, store, search, re-rank and connect to the LLM of your choice with a vision model trained on 1M pages of docs to turn your complex unstructured content into LLM-ready data.
Trusted by firms like like Samsung, Air France, Dartmouth and hundreds more who have ingested more than 3B tokens of data on the platform
Operate entirely within your network, including air-gapped environments, ensuring complete data sovereignty and compliance with internal security policies.
GroundX ensures robust protection with AES-256K encryption, securing your data both at rest and transit. With API key-based security as the sole access control mechanism, GroundX maintains a strict perimeter while minimizing attack vectors, delivering security by design with a minimal attack surface.
GroundX Ingest employes a unique process to turn complex unstructured documents into LLM ready data. We trained a vision model on nearly 1M pages of enterprise documents across many industries. The visio model identifies complex objects – tables, graphics, forms, schematics, text and more - on the page, then passes each object through an agentic pipeline powered by a fine tuned VLM (visual language model) that turns the objects into LLM ready data.
Built to handle a highly diverse set of documents for RAG including
The GroundX ingest models can also be fine tuned for your unique data set.
In head to head research, GroundX was 50% more accurate than other RAG frameworks, especially on complex visual documents. Read the study.
GroundX automatically balances workloads across available GPU and CPU resources to help you grow to almost any size without the setup and engineering.
GroundX is built on Kubernetes, fine-tuned open source models and optimized for Open Shift. Under the hood you'll find open source software you trust including OpenSearch, Reddis, Kafka, MySQL and MinIO.
Red Hat certified products are tested to meet Red Hat’s criteria and supported as defined in the Red Hat Collaborative Support Process.
Partner validated products are tested by Red Hat Partners and supported as defined in the Red Hat Third Party Component Policy.
GroundX is a feature complete RAG engine that handles ingest, process, store, search and re-rank, then hands retrievals to the LLM of your choice.
GroundX is compatible with all major completion models from closed source like OpenAI and Google Gemini to self-hosted open-source like Llama and Mistral.
We merged a vision model with a visual language model, then trained them on nearly 1M pages of enterprise documents from many industries. The result is SOTA document understanding on everything from medical bills to police reports to mechanical diagrams. And you can fine tune GroundX ingest to your unique documents.
You need both GPUs and CPUs to run GroundX ingest and search. Modest GPUs, like NVIDIA L40S, can ingest nearly 1M pages per month. An A100 can ingest more than 150M pages of RAG content a month. The specs are here.