Thatffener Neuro® to Multi-Agent Accelerator
Cognizant (Nasdaq: CTSH) announced today that it Neuro® AI Multi-Agent Accelerator for research and academic purposes as spring-open software. This Open-Source-Software It enables domain experts, researchers and developers to immediately start prototyping and build up agent networks for practically every application. The Open-Source-Software will help accelerate the introduction of AI by promoting the cooperation in the development and adaptation of multi-agent systems for adaptive operations and real-time decisions. Companies can use the Multi-Agent Services Suite from Cognizant to use agent networks in a commercial environment to a large extent and to efficiently manage them as part of a commercial license.
The Market for AI agents It is expected to grow rapidly in the next five years, of a value of $ 5.1 billion in 2024 to a forecast value of $ 47.1 billion by 2030. Today’s messages can illustrate how companies can use networked agents to open up new sources of income and generate scalable business value. The open source software for the Neuro® AI Multi-Agent Accelerator developed by the AI Lab of Cognizant underlines Cognizant’s leadership role in the field of AI innovation and its commitment to the further development of AI agents.
Customers such as Telstra, Australia’s leading telecommunications and technology companies, work with cognizant to test and use multi-agent systems.
“The open source publication of the Neuro Ai Multi-Agent Accelerator will further enable our teams, to quickly develop prototypes and to integrate existing AI agents, and help to accelerate our software development cycle,” said Kim Krogh Andersen, group leader for product and technology at Telstra. “We already recognize the potential for improvements in terms of quality, speed and efficiency.”
Cognizant also supported a healthcare company in setting up an agent network for contract negotiations that shorten the processing times for medical objections, as well as a consumer goods company in analyzing supply chain management. Cognizant currently has more than 65 conversations with customers about agent -based AI.
The establishment of a successful multi-agent network requires the ability to coordinate various agents, tools and sources of knowledge-including universal Large Language Models (LLMS) and organizational systems such as Service Level Management (SLMS) or Retrieval Augmented Generation (RAG) frameworks. The Neuro AI Multi-Agent Accelerator from Cognizant aims to enable almost seamless integration with APIs, RAG and third-party agent such as Agentforce from Salesforce, Agentpace from Google or Crew AI-via its native model Context Protocol (MCP) or standard API calls. An optional protocol for coordination between agents enables these agents to organize themselves independently, to delegate tasks and to lead processes, which increases efficiency and minimizes errors. The Agent2Agent (A2A)-Protocol is also supported, which expands the cooperation between agents across clouds, platforms and corporate borders.
“In order to remain competitive in the age of agent -based AI, companies must have freedom to experiment – to research how agents can transform business processes and increase operational efficiency,” said Babak Hodjat, head of the KI technology area at Cognizer. “With the open source release of the Neuro Ai Multi-Agent Accelerator, we expand access to our state-of-the-art multi-agent technology. This enables developers to enable faster innovations and transfer decisions regardless of their technical background to quickly create prototypes of systems and to observe their effects on important performance indicators directly.”
“Agentforce is based on the extensively integrated platform of Salesforce, which is open and expandable and offers our ecosystem of partners and developers the opportunity to advance with a trustworthy AI innovations. To provide the decision of Cognizant to provide its neuro Ai Multi-Agent Accelerator as an Open Source is an excellent example of the kind of partnership that helps our customers To get in advance and be innovative with confidence, “said Gary Lerhaupt, Vice President for Product Architecture at Salesforce. “Together we enable companies to use agents who think, work together and create added value in all areas of their company.”
Neuro® AI Multi-Agent Accelerator-Main features:
- Intelligent discovery of opportunities: Enter a company name or a problem area, and the Agent Network Designer automatically suggests a tailor -made agent network that is tailored to your application. So you get from the idea for implementation faster.
- Fast and optimized adjustment: Create and modify multi-agent systems quickly with the help of natural language or use prefabricated templates for areas such as lending, customer service, retail optimization and intranal automation to significantly reduce development cycles and risks.
- Scalable, distributed operation: The connectors support self-developed tools, APIs and third-party agents such as Agentforce from Salesforce and Google Agenspace via the Model Context Protocol (MCP) or standard API calls. A coordination layer enables agents to organize themselves intelligently, to distribute tasks and to guide processes, which increases efficiency and reduces errors.
- Safe private data: Supports regulated industries such as finance and healthcare through the isolation of confidential information about private data channels to ensure compliance and data protection.
- LLM and cloud provider independence: Simply switch between open source and most commercial LLMs as well as between private and public cloud providers without having to set up your system.
- Extensive encoded tools: Improve your agent networks with tailor-made tools-these are essential to rely on real-time data or define logical limits that require human intervention if necessary.
- Several servers, distributed provision: Enter agent subnetworks over several servers and enable scalable architectures that support parallel processing, geographical distribution or segmented application cases.
- Data -controlled network definition: Define and update your agent -based systems entirely via pure configuration files that support version control, verifiability and quick reusability across projects.
- Agent test function: Use the Agent Network tester to recognize bottlenecks or failures in your network. Get implementable knowledge about coordination problems, gaps in the logic of agents or integration errors.
Cognizant scales agents for 330,000 employees through his intranet, 1 cognizant
Thousands of employees now use 1 cognizant, an intranet assistant based on the Neuro Ai Multi-Agent Accelerator. This tool consolidates and organizes several agents to efficiently support employees in various tasks, from the reservation of meeting rooms and ordering taxis to processing inquiries such as moving or getting married. 1Cognizant can now offer employees immediately implementable advice and support, which increases efficiency and brings out internal silos.
You can find more information on our Landing Page or in our Blog.
Information on cognizant
Cognizant (Nasdaq-100: CTSH) develops modern companies. We support our customers in modernizing their technology, the redesign of their processes and the transformation of their experiences, so that they are always one step ahead in our rapidly changing world. Together we improve daily life.
You can find more information at www.cognizant.com or @cognizant.