Coforge Unveils 'Intent Engineering' Framework to Govern Enterprise AI Agents

Coforge has introduced what it calls an intent engineering framework for enterprises deploying AI agents, adding a governance layer on top of the prompt and context engineering practices many teams already use. The framework is meant to define what autonomous systems should optimise for, how their performance should be measured, and when they should defer to human oversight.

The company splits enterprise AI into three layers. Prompt engineering influences how a model responds; context engineering supplies the information a decision needs; intent engineering builds the governance structure around the decision itself. Coforge's contribution sits in that third layer, where objectives, constraints and measurable outcomes determine what an agent should pursue.

"Most enterprises have invested heavily in teaching AI what to say and what it should know," Vic Gupta, Executive Vice President at Coforge, said in a statement. "The harder challenge is deciding what the agent should optimise for when objectives conflict, risk increases, or exceptions occur."

According to the company, the framework includes outcome-based design principles, controls built into the development lifecycle, escalation pathways for high-risk scenarios, and success metrics tied to business outcomes rather than activity volume. Together these aim to make agent behaviour explainable, bounded and accountable.

The framework arrives as enterprises hand AI agents access to more business systems and tools. The article points to a recent ServiceNow incident in which an AI agent attempted to change employee salaries, underlining the risks that come with giving autonomous systems access to sensitive actions.

Coforge says the approach draws on its work on enterprise AI programmes. It is intended for situations where business objectives conflict, risk increases, or an agent hits an exception that requires human intervention, moving governance upstream from post-hoc fixes into the design process.

The release follows earlier work on the operational side. In September 2026, Coforge expanded its AgenticOps capabilities to address governance, agent drift, security exposure, tool integration and observability as enterprises move agentic AI beyond experimentation into production.

As vendors productise the idea, the competitive ground for AI agents may shift from raw model capability toward governance and operability — who sets the objectives, who defines the boundaries, and who is accountable when something goes wrong.