Skip to main content

What is Private GPT

info

Private GPT (Software as a Service) empowers a customer directly by jumpstarting with an easy to use (ChatGPT-like) product that runs NVIDIA-optimized Open-Source models with integrated privacy, security and guardrails. By leveraging Retrieval-Augmented Generation (RAG), Private GPT enables LLMs to access and utilize your corporate and proprietary knowledge efficiently, ensuring responses are both accurate and contextually relevant. The modular architecture allows customers to scale their solution by integrating data connectors and fine-tuning models to meet their specific needs. Nebul ensures optimal RAG implementation with robust data connectors while also offering expertise in training and fine-tuning industry-specific and custom models.

GenAI on your terms​

Private GPT framework

The Private GPT framework is a modular and scalable platform that allows you to build, train, and deploy custom AI models on your own infrastructure (Private instance on Nebul's NeoCloud). It provides a flexible and secure environment for experimenting, testing, and deploying AI use cases.

What do you get?​

Private GPT (Business or Enterprise)​

  • Private URL (name.chat.nebul.io)
  • Private GPT (Frontend)
  • Private GPT AI Engine (Backend)
  • Nemotron Foundation Model (flatfee use) or Open Source model
  • Single Sign On (e.g. KeyCloack)

Add-ons​

  • Private GPT security
    • On browser - protect your employee
    • On product - protect your AI
  • Adapters + RAG optimalisation - read more on Add-on: to Adapters (GPT)
    • The option of enabling unlimited data adapters.
  • Models + Fine-tuning - read more on Add-on: to Models (LLMs)
    • LLMs and SLMs