Nvidia AI Blueprint
Image Credit: Nvidia
Nvidia announced its Nvidia AI Blueprint, which will make it easier for developers to build AI agents that analyze video and images content. Nvidia says that with this technology, any industry can search and summarize vast volumes
of visual data. Accenture, Dell, and Lenovo are some of the companies that are using a new Nvidia AI Blueprint for developing visual AI agents. These agents can boost productivity, optimize process, and create safer spaces.
Companies and public sector organizations are developing AI agents around the world to boost the abilities of workers who rely on visual data from a growing variety of devices, including cameras, IoT sensor and vehicles.
A new Nvidia AI Blueprint will allow developers from virtually any industry to create visual AI agents for video and image analysis. These agents can provide answers to user questions, create summaries, and activate alerts in specific scenarios. The blueprint is part of Nvidia Metropolis – a set developer tools to build vision AI applications – and is a customizable workflow combining Nvidia computer imaging and generative AI. Accenture, Dell, and Lenovo, global systems integrators and technology solution providers, are bringing the Nvidia AI Blueprint to cities and businesses worldwide. This will kick-start the next wave of AI apps that can be deployed in factories, shops, airports and traffic intersections. The Nvidia AI Blueprint, announced ahead of the Smart City Expo World Congress in early 2019, provides visual computing developers with a complete suite of optimized software to build and deploy generative AI-powered agent that can ingest or understand massive volumes live video streams and data archives.
Users are able to customize these visual AI agents using natural language prompts, instead of rigid software codes. This lowers the barrier for deploying virtual assistants in industries and smart city applications.
Nvidia AI Blueprint harnesses models of vision and language
Visual AI is powered by VLMs, a class generative AI models which combine computer vision with language understanding in order to interpret the world and perform reasoning.
Nvidia’s AI Blueprint can be configured to use VLMs such as Nvidia VILA and LLMs such as Meta’s Llama 3.0 405B, and AI models that are GPU-accelerated for question answering and context aware retrieval-augmented generation. Developers can easily swap out other VLMs, LLMs, and graph databases using the Nvidia NeMo Platform for their unique environments and usage cases. The Nvidia AI Blueprint can save developers months in research and optimization of generative AI models.
When deployed on Nvidia GPUs in the cloud or on premises, it can greatly accelerate the process of combing video archives to identify key events.
An AI agent created with this workflow in a warehouse could alert workers when safety protocols are violated. An AI agent at busy intersections could identify traffic accidents and generate reports that would aid emergency response efforts. In the field of public infrastructure workers could ask AI agents review aerial footage to identify degrading tracks, roads or bridges in order to support proactive maintenance.
Beyond the smart spaces, visual AI agents can also be used to summarize video for people with impaired sight, automatically generate recaps and help label massive datasets.
This video search and summary workflow is part of a collection Nvidia AI Blueprints which makes it easy to create AI powered digital avatars, create virtual assistants for personalized service and extract enterprise insight from PDF data.
Nvidia AI Blueprints can be downloaded and experienced by developers for free. They can then be deployed on accelerated data centers or clouds using Nvidia AI Enterprise. This software platform accelerates data science pipelines, streamlines generative AI deployment, and is available to all developers.
AI agents deliver insights from warehouses and world capitals.
Enterprises and public sector clients can also leverage the full collection Nvidia AI Blueprints through Nvidia’s ecosystem of partners. Accenture, a global professional services company, has integrated Nvidia AI Blueprints in its Accenture AI Refinery. This is based on Nvidia AI Foundry. Customers can create custom AI models using enterprise data.
Global system integrators from Southeast Asia, including ITMAX Malaysia and FPT Vietnam, are developing AI agents using the video search and summary Nvidia AI Blueprint to support smart city and intelligent transport applications.
Developers are able to build and deploy Nvidia AI Blueprints using Nvidia AI platforms, which include compute, networking, and software from global server manufacturers. Nvidia AI Blueprints can be found in the Dell AI Factory, along with Nvidia Hybrid AI Solutions.
Companies such as K2K, an application provider for smart cities in the Nvidia Metropolis eco-system, will use the Nvidia AI Blueprint in order to build AI agents which analyze live traffic cameras real-time. This will allow city officials to receive feedback on street activity, and ask questions. The company is also working with city traffic controllers in Palermo (Italy) to deploy visual AI agents utilizing NIM microservices, and Nvidia AI Blueprints.
Nvidia is going to talk about this more at the Smart Cities Expo World Congress in Barcelona, which runs from November 7 through 11.
Want to impress your boss? VB Daily can help. We provide you with the inside scoop on what companies do with generative AI. From regulatory shifts to practical implementations, we give you the insights you need to maximize ROI.
Read our privacy policy
Thank you for subscribing. Click here to view more VB Newsletters.
An error occured.
