Unlike other IT companies that incorporate hundreds of AI agents into their services, Cognizant has chosen a different approach. The company is integrating ‘vibe coding’ as a key part of its broader AI-driven strategy.
“Vibe coding is emerging as a key catalyst in this journey, empowering everyone—from seasoned developers to those who have never written a line of code—to build, solve, and innovate with the power of AI at their fingertips,” a Cognizant spokesperson told AIM in an email response.
The idea is simple but radical. CEO Ravi Kumar S, in a conversation with Lovable CEO Anton Osika, said that anyone, not just software engineers, should be able to build software using natural language. And machines, not humans, should write most of the code.
“I’m a big believer that we could produce more by democratising it [technology], and I’m hoping my software development community gets more productive using levelling,” Kumar said in the podcast, when Osika asked him how the company is using Lovable in their latest vibe coding journey.
Cognizant aims to generate 50% of its code using AI within a year and states it has already achieved 20% of that target.
The company has already trained 35,000 developers on GitHub Copilot through its Synapse skilling program and plans to train an additional 40,000 by the end of 2025.
“At Cognizant, we are harnessing AI to reimagine our company and equipping our associates with AI fluency, a gateway to new capabilities, roles, and opportunities,” the company stated.
‘Vibe Coding is Serious Business’
In a recent blog, Cognizant discussed the Windsurf acquisition by Cognition and cited Andrej Karpathy, stating that by embracing the vibes and ignoring the code, it has made coding accessible to all.
But this isn’t just about making engineers more efficient. Cognizant is trying to change who gets to write software in the first place.
I’m hoping to introduce more non-STEM disciplines into software development cycles,” Kumar added. “We have too many engineers trying to write code for the world. I want people with a sense of diversity. I want an anthropologist, sociologist, or even a psychologist to code.”
The goal is to address the increasingly diverse, complex, and real-world problems that Kumar describes. “It’s all driven by software. We aim to be a pioneer in this area.”
Software should not be the exclusive domain of computer science graduates. With tools like Microsoft Copilot and Cognizant’s own open-source Neuro platform, even a liberal arts graduate should be able to build functional prototypes by telling an AI what they want.
To prove the point, Cognizant is launching a company-wide event called Vibe Coding Week on July 30. This isn’t a small side project. Cognizant is supporting this shift with a significant investment of $1 billion over three years to develop platforms, train staff, and open-source its tools.
All Cognizant associates will participate in hands-on workshops with AI coding assistants, explore a prompt engineering toolkit, join best-practice sessions, and participate in innovation competitions, engaging with AI experts and thought leaders.
What about Business?
One of the key components of Cognizant’s infrastructure is the previously mentioned Neuro AI Multi-Agent Accelerator. This framework helps teams build networks of AI agents to handle tasks ranging from code writing to testing, deployment, and even documentation.
Neuro was open-sourced in May 2025, and several developers have already forked the codebase.
In many ways, this approach is about increasing volume and speed. Cognizant aims to make software development more industrialised—producing more code, more quickly, with a larger team. Yet, there’s also a meaningful change happening beneath the surface.
Highlighting their recent research, Cognizant stated that business leaders recognise there is work to be done before fully capitalising on the vibe coding opportunity.
“Even while 94% of senior leaders feel they’d lose out on productivity gains if they didn’t adopt AI, only 43% said they have established formal policies, procedures and guidelines regarding its use,” read the blog.
But client deployments are already underway. Telstra is using Cognizant’s Neuro agents to prototype service workflows. A US-based healthcare payer is using AI agents to streamline contract negotiations, thereby reducing the turnaround time on medical appeals. Aker Solutions has signed a multi-year deal to modernise its IT systems using the same framework.
Accenture, Infosys, and TCS are all investing in GenAI, but they haven’t open-sourced agent frameworks or committed to challenging targets, such as 50% machine-generated code. Cognizant’s decision to go public with its roadmap puts pressure on the rest of the industry to follow suit.
If vibe coding proves effective at this scale, it won’t just transform the way software is developed, but it will also shift the demographics of who creates it.
The post Cognizant’s Vibe Coding Lesson for Indian IT appeared first on Analytics India Magazine.
