NVIDIA has launched NIM Blueprints, a comprehensive catalog of pre-trained, customizable AI workflows aimed at revolutionizing the enterprise segment. This suite is designed to facilitate the building and deployment of generative AI applications for various critical use cases, such as customer service avatars, retrieval-augmented generation, and virtual screening in drug discovery.
According to Jensen Huang, founder and CEO of NVIDIA, the rapid advancements in generative AI and the expansion of model capabilities have ushered in a new era for enterprise AI. The NVIDIA AI Enterprise toolkit, which includes NeMo, NIM microservices, and the latest NIM Agent Blueprints, empowers enterprises to tailor open-source models, build custom AI applications, and deploy them seamlessly across any cloud, on-premises infrastructure, or edge environments.
NIM Blueprints offer a significant advantage for developers by providing ready-made, sample applications built with NVIDIA’s advanced technologies including NeMo and NIM alongside partner microservices. These blueprints come with reference codes, customization documentation, and a Helm chart for streamlined deployment. They enable enterprises to customize the workflows using their unique business data, thereby allowing continuous improvement and refinement of AI applications based on user feedback.
Currently available NIM Blueprints include a digital human workflow for customer service, a generative virtual screening workflow for drug discovery, and a multimodal PDF data extraction workflow for enterprise retrieval-augmented generation (RAG). These blueprints can be experienced and downloaded at no cost and can be deployed in production environments using the NVIDIA AI Enterprise software platform.
Several global system integrators and technology solution providers, including Accenture, Deloitte, SoftServe, and World Wide Technology (WWT), are set to deploy NVIDIA NIM Blueprints worldwide. Companies like Cisco, Dell Technologies, Hewlett Packard Enterprise, and Lenovo are offering complete NVIDIA accelerated infrastructure and solutions to facilitate rapid deployment.
The digital human NIM Blueprint, which is critical for advancing customer service, allows enterprises to adopt 3D animated avatar interfaces for more engaging user interactions. This blueprint integrates technologies such as NVIDIA ACE, Omniverse RTX, Audio2Face, and Llama 3.1 NIM microservices to enhance existing generative AI applications.
For enterprises dealing with large amounts of PDF data, the multimodal PDF data extraction NIM Blueprint leverages NVIDIA NeMo Retriever NIM microservices. It allows for the creation of digital humans, AI agents, and chatbots that can provide expert-level responses based on extensive datasets.
The generative virtual screening NIM Blueprint for drug discovery significantly speeds up the identification and optimization of promising drug-like molecules. It incorporates BioNeMo NIM microservices, including tools like AlphaFold2, MolMIM, and DiffDock, to expedite the virtual screening process using generative models.
NVIDIA plans to release additional NIM Blueprints monthly, catering to various applications in customer experience, content generation, software engineering, and product R&D. Enterprises will also have the ability to connect multiple Blueprints to develop more sophisticated and specialized generative AI applications.
NVIDIA’s extensive partner ecosystem is well-prepared to support enterprises in building and deploying generative AI applications using NIM Blueprints. Accenture, for example, is incorporating NIM Blueprints into its AI Refinery to further enhance its service offerings.
