Cognizant announced advancements built on NVIDIA AI aimed at accelerating the cross-industry adoption of AI technology in five key areas: enterprise AI agents, industry-specific large language models (LLMs), digital twins for smart manufacturing, foundational infrastructure for AI, and the capabilities of Cognizant’s Neuro® AI platform to integrate NVIDIA AI technology and orchestrate across the enterprise technology stack.
Cognizant is working with global clients to help them scale AI value efficiently, leveraging extensive industry experience and a comprehensive AI ecosystem comprising infrastructure, data, models, and agent development powered by proprietary platforms and accelerators. NVIDIA AI plays a key role in Cognizant’s AI offerings, with active client engagements underway across industries to enable growth and business transformation.
At NVIDIA GTC 2025, Cognizant presented its intent to deliver offering updates across the following five areas:
Enterprise AI agentification powered by Cognizant® Neuro AI Multi-Agent Accelerator: Running on NVIDIA NIM™ microservices, this framework will enable clients to rapidly build and scale multi-agent AI systems for adaptive operations, real-time decision-making and personalized customer experiences. With these frameworks clients can create and orchestrate agents using a low-code framework or use pre-built agent networks for various enterprise functions and industry-specific processes such as sales, marketing, and supply chain management. The frameworks also allow clients to easily integrate third-party agent networks and most LLMs.
Building multi agents for scale: Cognizant works to enhance business operations through the use of multi-agent systems and integration with NVIDIA NIM, NVIDIA Blueprints, and NVIDIA Riva speech AI. The company will be developing a future-proof agent architecture that supports modular and adaptable agent design to meet evolving needs and the long-term viability and adaptability of AI solutions. This includes pre-built integrations with security guardrails and human oversight. This approach aims to enable enterprises to develop and deploy market-ready applications tailored to their specific needs using the pre-built agent catalog. Examples include industry agents such as insurance claims underwriting multi-agent systems, appeals and grievances multi-agent systems, automated supply chain multi-agent systems and contract management multi-agent systems.
Industry LLMs: Cognizant is developing industry-oriented LLMs powered by NVIDIA NeMo and NVIDIA NIM. These solutions are tailored to meet the unique needs of different industries and build on Cognizant’s deep industry expertise to drive innovation and improve business outcomes. For example, Cognizant has developed a fine-tuned language model to transform healthcare administrative processes. This system will leverage Cognizant’s domain expertise and NVIDIA technology to enhance medical code extraction and support higher accuracy, reduced errors, and better compliance with HIPAA and GDPR standards. It is designed to help clients cut costs, decrease latency, improve revenue cycle management and help ensure accurate risk adjustment. In internal Cognizant benchmarking, the model has demonstrated effectiveness in reducing effort by 30-75 percent, boosting coding accuracy by 30-40 percent, and accelerating time to market by 40-45 percent.
Infrastructure for AI: Implementing AI effectively requires robust AI infrastructure and data prepared for AI. Cognizant’s infrastructure for AI, accelerated by NVIDIA, will provide clients access to NVIDIA AI technology via “GPU as a Service”, along with secure and managed infrastructure. This helps ensure that AI models can be run in various environments, including the cloud, data centers or at the edge. Additionally, Cognizant intends to use NVIDIA RAPIDS™ Accelerator for Apache Spark to help clients accelerate data pipelines for AI implementations, facilitating efficient and scalable operations. In one example implementation for a large healthcare client in the U.S., use of Cognizant’s infrastructure for AI resulted in a 2.7x cost efficiency improvement and a 1.8x enhancement in the performance of their Spark workloads.
Annadurai Elango, President, Core Technologies and Insights, Cognizant
We continue to see businesses navigating the transition from proofs of concept to larger-scale implementations of enterprise AI. Through our collaboration with NVIDIA, Cognizant will be building and deploying solutions that accelerate this process and scale AI value faster for clients through integration of foundational AI elements, platforms and solutions.
Jay Puri, Executive Vice President of Worldwide Field Operations, NVIDIA
From models to applications, enterprise AI transformation requires full-stack software and infrastructure with access to domain-specific data. The Cognizant Neuro AI platform is built with NVIDIA AI to deliver specialized LLMs and applications to ready businesses for the era of AI with reasoning agents and digital twins.
Nitish Mittal, Partner, Everest Group
As we enter the era of AI industrialization, enterprises are seeking to accelerate the value velocity of their AI investments—focusing on outsized economic impact, agentic-led workflow transformation, and industry-specific deployments. Cognizant’s deepening partnership with NVIDIA signals the right trajectory for forward-thinking enterprises aiming to unlock breakthrough value in the AI era.