Consider a scenario in which software development teams do not merely use AI for the writing of code but are able to coordinate AI agents throughout the process. This is the kind of change Coforge is after with the ASF.
Enterprises are looking at how to incorporate AI into their software engineering and find that it is not enough to simply have the code generated. There is the matter of making sure the software is tested, validated, reviewed and delivered in a secure and consistent manner that accords with enterprise standards. The ASF has been put in place to handle this with an operating model that is structured to manage AI agents over the whole of the SDLC. In so doing, it takes companies past developer copilots and early forays with AI to a more scaled form of AI-native engineering.
How Coforge’s Autonomous Software Factory Coordinates AI Agents
The ASF offers a way to govern, execute and orchestrate AI agents with human oversight as part of the equation. Operating in secure environments, these specialised agents can carry out quality assurance, testing, validation and code review in parallel with coding. It is a repeatable and scalable method for AI-native software engineering.
For the enterprise team the goal is to get beyond individual tasks and see AI coordinated across several stages of the engineering lifecycle. “The future of software engineering is in agent orchestration,” is the view at Coforge.
Vic Gupta, Executive Vice President of Coforge’s Microsoft Business Unit, puts it like this: “Enterprise software development is in a post-coding era. Generating the software is becoming the easy part. Where organisations will gain a competitive edge is in how well they can coordinate, govern and scale AI agents over the course of the software lifecycle.”
In such a model, the AI agents do an increasing amount of the execution while the humans are left to deal with architecture, strategy and business results.
Why AI-Native Software Development Needs Governance and Human Oversight
Quality, compliance and control need to be rethought by enterprises as the output from AI makes up a larger share of development. While the ASF has the necessary governance mechanisms and human oversight, Coforge has not put any specifics on implementation or customer deployments in its announcement, nor has it quantified any productivity gains.
This is a point of note for businesses considering agentic software development. To make room for AI agents one has to do more than automate; it requires a redesign of the engineering process and the models of governance and quality control.
Coforge’s AI-Native Engineering Strategy and Industry Context
Coforge is building on the AI Adoption Fabric it put forward in August 2026 to clear the barriers of governance and change management for agentic development. That involved getting the workforce ready and overhauling the controls for an agentic SDLC.
The move is in step with other Indian IT services firms putting out their own agentic AI. HCLTech is working Claude and Claude Code from Anthropic into its AI Force for IT support and code remediation. TCS has the Industrial Intelligence Foundry for factory floor AI agents and LTM has opened a lab in Chennai for industrial AI and robotics.
Coforge pairs its AI agents with an AI-enabled workforce, including Forward Deployed Engineers in hybrid pods. The message to the enterprise is that having the right tools is only part of what is needed for AI-native software engineering. It is a question of coordinating the agents and seeing that delivery is secure and governed. With its emphasis on those things, Coforge is making a play for a time when the management and construction of software is of equal import to the code.
