Cognizant puts 25,000+ people on Claude delivery and will run Claude systems in production for clients

Cognizant announced on November 10 a Claude Frontier Capability Group: more than 25,000 professionals in a dedicated global organization that builds and runs Claude-powered systems inside client businesses. The company says more than 44,000 associates are already trained or certified on Claude and that it holds more than 26,000 Claude certifications. It also launched a managed services model in which Cognizant, rather than the client, operates Claude-powered systems in production, bundling staff, licensed Claude access and its own IP.

The group starts with four focus areas: redesigning business domains around AI, building and modernizing the supporting systems (including Claude agents for legacy and mainframe code), operating agents securely at scale, and AI-native software engineering. Cognizant cited client results such as shop-floor agents at a pharmaceutical manufacturer that cut downtime by 35%. Both initiatives rely on its Frontier workforce model, which trains Frontier Certified Engineers and Frontier Business Operators credentialed by the model companies.

Why it matters for FDEs: this follows Cognizant's 18,000-person Gemini group (October 19) and its 25,000-person OpenAI group (October 30). The big IT services firms are now organizing deployment staff by model vendor and selling the hands-on, in-the-client work that forward deployed engineers do, at very large scale. For engineers, vendor certifications and production deployment experience on a specific frontier model are becoming the entry ticket to these teams, and the managed services angle means more of the work continues after go-live.