The company says its new human-capital model is designed to turn enterprise AI spending into measurable business outcomes, with the first deployable cohort expected by the fourth quarter of 2026.
Cognizant said it plans to expand its Frontier-certified workforce to 5,000 engineers and 10,000 operational leaders as it builds what it describes as the human and operating infrastructure needed to convert AI capabilities into measurable business results. The company said its first class trained under the Frontier method will be assessed and ready for deployment by the fourth quarter of 2026, while annual hiring from U.S. and international universities will help deepen the talent pipeline. The company framed the move as a response to what it called a gap between heavy corporate AI spending and limited realized returns, estimating that difference at $4.5 trillion. Cognizant said the shortfall is primarily a people-and-process issue rather than a computing-capacity problem, and that simply adding more infrastructure will not solve it. Chief Executive Officer Ravi Kumar S. said Frontier workers are intended to combine deep client-industry knowledge with the ability to redesign how work is organized and take responsibility for outcomes across whichever model or cloud a client uses. Cognizant said the platform is model-agnostic and cloud-agnostic, and works alongside technology already chosen by clients through partnerships that include Anthropic, OpenAI, Microsoft, Google, AWS, NVIDIA, Salesforce and ServiceNow. The company said the model rests on six principles, including interdisciplinary work, direct linkage to client value, daily creation or collaboration with agents, end-to-end accountability, small operational teams and a unified client experience. It divides roles into two tracks: Frontier engineers, who build and operate multi-agent systems (multiple AI agents working together), and Frontier operational leaders, who manage outcomes across mixed human-and-digital workforces. Cognizant said the framework is already being used. In one recent project for a large restaurant-sector company, a team of one engineer and one operator redesigned account-management work using 17 production AI agents, freeing about 11 hours a week for each account manager, cutting handoff cycles by about 60% and nearly tripling revenue per interaction. To support the effort, Cognizant said it is broadening SkillSpring and expanding AI literacy and responsible AI training. The company said structured AI-Bridge programs have already enabled 40,000 employees to gain Frontier certification directly from companies behind tools including GitHub Copilot, Google Gemini, Claude from Anthropic and Codex from OpenAI. Cognizant said the broader aim is to help clients generate financial returns from AI investments by pairing technology with durable operating accountability.