Cognizant (NASDAQ: CTSH) announced plans to scale its AI-focused workforce to 5,000 Frontier Certified Engineers and 10,000 Frontier Business Operators, with the first deployment-ready cohort expected by the fourth quarter of 2026.

The company said it will also hire Frontier-native talent annually from U.S. and international universities to supplement its internal training pipeline. The initiative will expand Cognizant's SkillSpring training platform and fund embedded client engagements worldwide.

Cognizant described the two workforce tracks as distinct roles. Frontier Certified Engineers are responsible for building and deploying agentic AI systems and maintaining accountability after go-live. Frontier Business Operators manage workflows in environments where human and AI agent teams operate together, drawing on operational experience to refine AI systems over time.

The company cited a $4.5 trillion gap between what AI can deliver and what enterprises currently realize, characterizing the shortfall as a people and process problem rather than a technology capacity issue.

CEO Ravi Kumar S. said the model focuses on outcome accountability rather than technology deployment alone. "By taking accountability for outcomes rather than stopping at technology deployment, we can help clients accelerate measurable results while managing risk," he said.

Cognizant said the workforce model is cloud- and model-agnostic, with partnerships spanning Anthropic, OpenAI, Microsoft, Google, AWS, Nvidia, Salesforce, and ServiceNow.

The company provided one deployment example: a two-person engineer-and-operator pod that converted a food service company's account-management workflow into 17 production AI agents, reclaiming roughly 11 hours per account manager weekly, reducing handoff cycles by approximately 60%, and nearly tripling revenue per engagement.

The broader certification pipeline targets 40,000 associates across Frontier programs, credentialed by frontier-model companies including GitHub Copilot, Google Gemini, Anthropic's Claude, and OpenAI's Codex.