Cognizant’s Neuro AI Accelerator is now available as open source. With this move, Cognizant aims to lower the barrier for AI adoption.
Cognizant is making its launched earlier this year Neuro AI Multi-Agent Accelerator open source. This allows developers, researchers, and companies to experiment with scalable AI agent networks for various applications. The move is intended to accelerate the adoption of agentic AI in the business world.
Flexible Building and Testing
The Neuro AI Multi-Agent Accelerator is a platform for building and managing collaborative AI agents. The software supports integration with existing systems via APIs, large language models with retrieval augmented generation (RAG), and well-known platforms such as Salesforce, Google, and ServiceNow.
Thanks to the Model Context Protocol and the optional coordination protocol, agents can independently distribute and execute tasks. Collaboration between agents across different cloud environments is also possible via the Agent2Agent protocol. This protocol was conceived by Google, but increasingly more companies are starting to embrace it so that AI agents from different vendors can communicate with each other.
The Neuro AI Accelerator platform was developed by Cognizant’s AI Labs. Companies can use the Multi-Agent Services Suite for commercial implementations, allowing these networks to be deployed and managed at scale.
Practical Applications
Cognizant mentions several examples where it deploys multi-agent systems. For instance, it developed a network of agents for a healthcare institution to handle contract negotiations, which reduced the processing time of medical objections. In the FMCG sector, the system was used for supply chain analysis.
Internally, Cognizant uses the technology for 1Cognizant, a chatbot that bundles multiple agents. Employees use this tool for various administrative and practical tasks. The chatbot replaces an older system and proves to work more efficiently.
According to Cognizant, the market for AI agents will grow significantly in the coming years, and the company is certainly not alone in this vision. By making the technology open source, the company wants to give businesses and researchers more room to experiment and develop tailored solutions. At the same time, commercial licenses remain available for organizations that want to deploy these systems on a large scale.