Cognizant Open-Sources Its Neuro AI Multi-Agent Accelerator
Global solution provider Cognizant has announced the open-sourcing of its Neuro AI Multi-Agent Accelerator technology, designed for prototyping and building agent networks across various industry verticals. This move aims to facilitate the development of multi-agentic systems and encourage collaboration within the AI and developer community.
The Neuro AI Multi-Agent Accelerator serves as the platform and framework used by Cognizant to construct multi-agent systems securely and at scale. According to Babak Hodjat, the Chief Technology Officer of AI at Cognizant, the decision to make this technology open source was driven by the lack of existing platforms matching the scale, interoperability, simplicity, and safety features offered by their solution.
Key Features and Benefits
The repository for Cognizant's multi-agent accelerator, known as neuro-san, will be hosted on GitHub. Additionally, a studio named neuro-san-studio will provide users with pre-built templates, including an agent network capable of creating further agent networks based on specific tasks or processes.
The Neuro AI Multi-Agent Accelerator is available under an academic research license, allowing developers and researchers to experiment with the technology and contribute to its enhancement. While commercial use is restricted under the license, the platform supports interoperability with various agent networks, including Google's Agentspace and Salesforce's Agentforce.
Interoperability and Use Cases
With support for Anthropic's MCP and Google's Agent2Agent, the Neuro AI Multi-Agent Accelerator offers a high level of interoperability, enabling seamless communication between different agent networks. Cognizant has leveraged the technology internally, integrating it into their intranet system, which includes agents from departments such as HR, finance, legal, and sales.
Furthermore, the system allows users to run multiple agent networks across different servers, facilitating complex interactions between agents. Cognizant emphasizes the scalability, security, and extensibility of the Neuro AI Multi-Agent Accelerator, highlighting its practical applications for modern enterprises.
Commercial Adoption and Future Prospects
Although Cognizant already has commercial clients utilizing the Neuro AI Multi-Agent Accelerator, the company envisions broader adoption within the industry. By making the technology open source, Cognizant aims to set a standard for creating scaled multi-agentic systems and foster innovation in AI development.
As the AI and developer community embraces this initiative, Cognizant expects to see collaborative efforts in advancing the capabilities of agent networks and driving the evolution of multi-agent systems across diverse sectors.
