The companies announced the expansion on October 19, 2026. The group will focus on moving enterprise AI agents from pilots into production.
Key figures:
- 18,000+ Cognizant associates in the new Gemini-focused group
- 55,000+ employees completed Build with Gemini training
- 100,000 Cognizant employees will receive Gemini Enterprise in 2026
- Internal deployment is planned to expand to 200,000 employees
Cognizant Builds an 18,000-Person Gemini Group
The group will develop and deploy agentic AI systems that can perform multi-step tasks, access enterprise data and interact with business applications.
Cognizant engineers will use Google Cloud’s Gemini stack to build agents, connect them to corporate systems and deploy them in customer environments.
Gemini Enterprise: Core Technologies
| Component | Role |
| Gemini Enterprise | Interface for employees to access AI agents and enterprise data |
| Gemini Enterprise Agent Platform | Platform for building, deploying and managing agents |
| Gemini models | Foundation models for reasoning and generation |
| Model Garden | Access to 200+ foundation models |
| Agent Development Kit | Code-first framework for custom and multi-agent applications |
| Agent Studio | Low-code agent development |
| Agent Runtime | Managed execution environment |
| Memory Bank | Persistent agent memory |
| RAG Engine | Connects models with private enterprise data |
| Vector Search | Semantic and vector retrieval |
| Agent Registry | Inventory of agents, tools and MCP servers |
| Agent Identity | Individual identity and access controls for agents |
| Agent Gateway | Controls agent-to-agent and agent-to-tool traffic |
| Model Armor | Prompt-injection and data-leakage protection |
| Observability | Logs, metrics and traces |
| Protocols | MCP, REST, gRPC and Agent2Agent |
The security layer is designed for production use. Individual agents can receive separate identities and permissions, while Agent Gateway and Model Armor control access and inspect traffic.
How the New Deal Compares With the Earlier Partnership
| Earlier phase | October 2026 expansion | |
| Structure | Gemini Enterprise Practice | Gemini Frontier Capability Group |
| Workforce | Certified and forward-deployed engineers | 18,000+ associates |
| Main goal | Gemini deployment | Production-scale AI agents |
| Training | Gemini certification programs | 55,000+ trained employees |
| Internal rollout | 100,000-employee target | 100,000 in 2026, later 200,000 |
| Focus | AI implementation | Agent deployment, governance and operations |
Cognizant launched a dedicated Gemini Enterprise Practice earlier in 2026. The October expansion increases the dedicated workforce and shifts the focus toward production-scale agentic systems.
From Pilots to Production
The technical challenge is not access to an AI model but integrating agents with enterprise systems.
A production deployment typically requires:
- Gemini models for reasoning and generation
- RAG and connectors for private company data
- Agent Runtime and ADK for workflow execution
- IAM and Agent Identity for access control
- Agent Gateway and Model Armor for security
- Observability for monitoring and auditing
Cognizant and Google Cloud cited one retailer using Gemini across its software-development lifecycle.
According to the companies, the deployment produced:
- 40% higher developer productivity
- 30% shorter time-to-market
- 35% cost savings
These figures are company-reported. Cognizant has also reported internal tests showing up to 30% higher development velocity and automation of roughly 60–70% of manual work in selected workflows.
What Changed
The announcement does not introduce a new Gemini foundation model. The change is operational scale.
Google provides the AI models, agent platform, infrastructure and security tools. Cognizant provides the engineering workforce to integrate those systems into enterprise environments.
The central figure in the deal is therefore 18,000 specialists, not a new model benchmark. The partnership is moving from Gemini adoption projects toward large-scale deployment of production AI agents.
