Nvidia has launched a new artificial intelligence (AI) Blueprint aimed at helping organizations and public agencies develop visual AI agents that analyze video and image data.
This framework, part of Nvidia’s Metropolis toolset for vision AI applications, supports a wide range of industries by enabling developers to create agents capable of searching, summarizing, and providing alerts based on visual information from devices like cameras, IoT sensors, and vehicles.
Announced ahead of the Smart City Expo World Congress taking place in Barcelona through Nov. 7, the Nvidia AI Blueprint for video search and summarization gives developers a comprehensive set of tools to create customized AI agents that handle large volumes of video and image content. The agents can be tailored with natural language prompts, allowing users to avoid complex coding. This flexibility lowers the technical barriers for deploying AI agents in areas such as smart cities, transportation, and industrial sites.
The blueprint integrates Nvidia’s computer vision and generative AI technologies and includes access to various Nvidia NIM microservices and vision language models (VLMs), which combine visual data interpretation with language processing capabilities. This combination enables the AI agents to perform tasks such as identifying patterns in traffic footage or detecting maintenance needs in public infrastructure through video analysis.
Global integrators
Major tech companies, including Accenture, Dell, and Lenovo, are leveraging the Nvidia AI Blueprint to bring AI-enhanced solutions to different markets, enhancing efficiency and safety in locations like airports, factories, and busy intersections. For instance, an AI agent in a warehouse could alert workers when safety protocols are not followed, or in a smart city, the AI agent could analyze live traffic video to detect collisions and aid emergency responders.
Nvidia has collaborated with global integrators and technology providers, including ITMAX in Malaysia and FPT in Vietnam, to deploy these solutions across Southeast Asia. Nvidia’s partners, such as K2K, are applying the blueprint to monitor live traffic cameras, allowing city managers to respond to activity in real time. Nvidia AI Blueprints are available for download and can be implemented across various data centers and cloud systems through Nvidia’s AI Enterprise platform.
Nvidia’s partners also include Dell, which plans to integrate the blueprint’s capabilities into its AI Factory and NativeEdge platforms, enabling more edge AI applications. Similarly, Lenovo incorporates Nvidia AI Blueprints within its Hybrid AI solutions.