Cognizant and Google Cloud launch Gemini Frontier Capability Group
Cognizant and Google Cloud have expanded their partnership and created a Gemini Frontier Capability Group, supported by more than 18,000 dedicated Cognizant employees, to speed up the rollout of agentic AI in large enterprises. According to the companies, more than 55,000 Cognizant staff have also completed the largest Gemini partner workshop to date.
- Cognizant and Google Cloud created the Gemini Frontier Capability Group, backed by more than 18,000 dedicated Cognizant employees.
- More than 55,000 Cognizant staff completed the "Build with Gemini" workshop, described as the largest Gemini partner workshop so far.
- The group plans an MCP-based agent architecture secured by Google Cloud's Model Armor.
- A retailer reportedly saw 40% higher developer productivity, 30% faster time to market and 35% lower costs, according to the companies.
- Cognizant will give Gemini access to 100,000 employees this year and plans to reach 200,000.
A dedicated group for enterprise AI agents
Cognizant (Nasdaq: CTSH) and Google Cloud said the new group brings together a dedicated international team. It combines Google's end-to-end, enterprise-grade AI platform with Cognizant's domain, industry and engineering expertise as a builder of AI solutions. The companies describe it as a significant joint investment that builds on the momentum of Cognizant's Gemini competence area, set up in April 2026.
According to the announcement, the group is meant to ease the move from pilot projects to production. The aim is to let companies deploy robust, enterprise-grade AI agents securely, with measurable business value.
Ravi Kumar S, CEO of Cognizant, said that the group lets the company's best engineers focus fully on bringing the latest-generation AI into production for customers. Thomas Kurian, CEO of Google Cloud, said that deploying AI agents at scale requires trusted partners with deep expertise in systems architecture and integration.
Focus areas and technical approach
The group will initially concentrate its engineering skills on four areas:
- modernizing existing systems
- improving the developer experience
- moving business processes to autonomous AI agents
- advanced cyber defense
To make these capabilities usable in practice, the group plans to use a standardized, security-oriented architecture for AI agents. It is based on the Model Context Protocol (MCP) and further protected by Google Cloud's Model Armor.
These frameworks are intended to underpin a new set of collaborative, production-ready enterprise solutions. The companies name three examples: Cognizant's Catalog Support Agent (CSA) for retail, a Healthcare Agentic Marketplace orchestrated with Neuro San, and multi-agent Order-to-Cash (O2C) workflows for manufacturing.
Training and a first customer result
The partnership rests on scaling customer projects with Cognizant's Frontier Certified Engineers. As part of its certification drive, Cognizant recently completed the "Build with Gemini" workshop with Google Cloud. Participants built AI agents and applications and earned Gemini credentials. More than 55,000 Cognizant employees took part, which the companies call the largest single Gemini partner workshop so far.
As a first practical example, the companies say Cognizant integrated Gemini models across the whole software development lifecycle for a leading retailer. According to the announcement, this raised developer productivity by 40%, cut time to market by 30% and lowered costs by 35%.
Cognizant is also using Gemini internally at scale. As announced in July, it will make Gemini available to 100,000 employees this year and plans to raise that number to 200,000. It says it continues to expand its commitment to certifying professionals on the platform.
What it means for enterprise customers
For companies stuck between AI experiments and live deployment, the announcement signals a larger pool of certified engineers focused on Google's Gemini stack. The companies say demand for scaling AI in enterprises is rising, and the new group is their answer. The results cited so far come from a single retail customer and are the companies' own claims. Further customer outcomes will show how far the approach carries across the retail, healthcare and manufacturing solutions named in the release.
