Meet the AI-augmented engineers at Freshworks

Freshworks CEO Dennis Woodside wrote recently about how our engineering teams are putting AI to work, and showing some amazing results: The teams reduced coding time by 30% and  improved code quality by 61%.

The productivity gains were impressive, but they only tell part of the story. Engineers’ lives were changing—they were gaining new “coworkers” in the form of AI agents and copilots. The change began reshaping workflows and collaboration in unexpected ways. For engineer Aarthi Kathirvel and others, AI wasn’t just a productivity booster—it fundamentally changed how she learned, built, and contributed to projects.

To understand this transformation on a personal level, we spent time getting to know several developers and leaders on the team. What we found wasn’t just a shift in efficiency—it was a new vision for how engineers approach problem-solving, skill development, and teamwork.

Collaborating with AI

For Jeykar Watson, VP of engineering at Freshworks, AI offered a way to help his team balance speed and quality. Engineers spend significant time on repetitive tasks—debugging, refactoring, and maintaining legacy code. Watson saw an opportunity to shift that burden and free them up for higher-value work.

“AI tools weren’t about replacing developers—they were about enabling them,” says Watson. “We wanted engineers to focus on solving problems, not just writing boilerplate code.”

To ensure a smooth transition, Watson’s team started with a pilot program of 15 engineers, carefully tracking how they used AI tools in real-world scenarios. Some engineers immediately saw AI’s potential, integrating it into their workflows and championing its benefits. In addition to using AI to generate code, these engineers were exploring how AI could optimize workflows, improve debugging, and even suggest better architectural decisions. Others took more time to adapt, testing AI’s capabilities in smaller ways, using AI to improve coding workflows by recommending snippets and even writing large portions of code, leaving engineers to focus on reviewing and revising it.

“Those early adopters became our enablers,” says Watson. “They weren’t just using AI to generate code; they were exploring how it could optimize workflows, improve debugging, and even influence architectural decisions.”

As adoption grew, Watson established focus groups based on expertise areas—mobile development, backend engineering, and testing—so engineers solving similar challenges could learn from each other’s AI-assisted experiences. These groups documented their learnings, tracking efficiency gains and best practices.

We wanted engineers to focus on solving problems, not just writing boilerplate code.

Jeykar Watson

VP of engineering at Freshworks

The results were striking. One team reduced an estimated 70-hour project down to just eight hours by leveraging AI-generated code suggestions. Beyond time savings, AI changed how engineers approached problem-solving.

“Engineers weren’t just writing code—they were brainstorming better solutions and collaborating more creatively,” Watson explains. “That was the real win.”

Developing new skills

Engineers have been coding on their own for decades, so working in partnership with AI required some support and new skills. Varagunapandian (“VGP”), director of engineering at Freshworks, recognized that to get the most out of AI, engineers needed structured support and a fresh perspective on their roles.

“We viewed AI as an augmentation tool, not a replacement,” says VGP. “Our goal was to create an ecosystem where engineers felt empowered, not threatened.”

To facilitate this transition, VGP’s team launched hands-on workshops, live demos, and Slack-based learning communities where engineers could share AI-generated solutions in real time. Instead of traditional lectures, they introduced interactive coding sessions, allowing engineers to experiment with AI in live scenarios.

By rolling out AI incrementally and allowing engineers to discover its value for themselves, the program avoided top-down mandates and fostered organic adoption.

“More than just implementing AI, it was about helping engineers see AI as a collaborator, a teammate,” says VGP. 

Getting time back in the day

Before AI, Aarthi Kathirvel, a 26-year-old software engineer, spent a significant amount of time searching for syntax, combing through Stack Overflow, and piecing together best practices from scattered documentation. When tasked with a project using Tailwind CSS—a framework she had never worked with before—she expected to spend days learning the syntax and debugging errors.

Instead, she turned to GitHub Copilot. Rather than manually looking up styling conventions, she simply described the UI she wanted to build, and Copilot generated the corresponding Tailwind CSS code in seconds.

Read also: How Freshworks wins with Freshservice

“Instead of Googling ‘How do I add a 1px border in Tailwind?’ I could just describe the mockup, and it would generate the entire snippet for me,” Aarthi explains. “Google couldn’t do that.”

Beyond syntax, AI helped Aarthi navigate unfamiliar codebases. For a side project integrating two chatbot systems within Freshbots, she typically would have spent hours reading documentation or asking colleagues for explanations. Instead, Copilot analyzed the code and provided plain-language summaries of key methods and functions.

“It was like having an always-available mentor,” she says. “I didn’t have to interrupt someone’s day to ask, ‘Hey, what does this method do?’ I could just ask Copilot, get an explanation, and move forward.”

Now, what would have taken her three days of manual research and coding was reduced to a single day. More than just efficiency, AI gave Aarthi the confidence to take on projects she might have otherwise avoided.

“I had extra time I wouldn’t have had before, and that made me think: What else can I build?” she says.

AI-powered workplaces

AI has undeniably changed the way engineers at Freshworks work, learn, and collaborate. Some changes are practical—faster coding, streamlined debugging, and improved efficiency. Others are more profound—AI is shifting how engineers think, solve problems, and explore new ideas.

While AI can generate code, detect errors, and suggest optimizations, engineers still bring something AI can’t—creativity, judgment, and deep technical insight. AI is not replacing engineers; it is amplifying them.

“AI is not about doing the work for us,” says Watson. “It’s about freeing us to focus on the work that matters most.”