In July 2023, Suumit Shah, the founder and CEO of Indian e-commerce startup Dukaan, announced that his company had replaced 90% of its customer support team with an AI-powered chatbot. The announcement, made on social media, caused an uproar. While some observers praised the operational improvements, many others condemned the decision as tone-deaf and devoid of compassion.
The move sparked debate beyond the startup’s small Bengaluru office. It raised difficult questions about the balance between technological efficiency and human livelihood, and about the way companies communicate sweeping changes that directly affect people’s jobs and lives.
What Happened at Dukaan?
Dukaan is a platform that helps small businesses set up online stores. In the summer of 2023, Shah posted on Twitter (now X) that the company had replaced nearly all of its customer support team with a custom-built chatbot named Lina.
According to Shah, the change led to massive gains: customer wait times dropped from over 1 minute to zero, resolution time shrank from more than 2 hours to just over 3 minutes, and support costs were reduced by 85%. He emphasized that the transition took only two days and that the AI assistant could respond to both general inquiries and user-specific issues.
However, what may have been intended as a case study in operational streamlining quickly became a flashpoint. His tweet was flooded with criticism from users who felt the post lacked basic empathy. Many took issue not only with the layoffs but also with the seemingly celebratory tone in which they were announced.
Public Reaction
Online responses varied in tone but were largely negative. Critics questioned the wisdom of announcing layoffs as a technological milestone. Many accused the CEO of treating human workers as disposable and voiced concern that decisions of this kind were becoming more common without sufficient public discussion or oversight.
One Twitter user wrote, “What kind of company brags about firing its employees?” Another replied, “Automation is inevitable, but decency is a choice.”
Others defended the move on the grounds of efficiency. Some entrepreneurs and investors commented that startups must optimize every part of their operation, and that this step—however difficult—was ultimately logical given the cost of human support.
Still, the prevailing sentiment was one of discomfort. The conversation quickly expanded from a single company to a broader discussion about how firms are implementing new technology and what obligations they owe to their workforce during transitions.
A Sign of What’s to Come?
The decision at Dukaan wasn’t isolated. In the months following the announcement, multiple companies across various sectors began to experiment with generative AI tools, particularly in customer service, logistics, and content production.
Retail chains, financial institutions, and software platforms have each explored using AI in places where workers once handled routine tasks. Many of these systems now assist with or fully manage email replies, chat-based help desks, and call center conversations.
Executives from larger corporations have made similar statements. Microsoft, for example, has integrated AI models into its suite of office products, aiming to reduce time spent on administrative tasks. Wells Fargo and other financial firms have publicly acknowledged that automation is now a major part of their operational planning.
The Argument for AI in Support Roles
There’s a clear business case for introducing AI into support departments. Traditional call centers are expensive to operate, often require large teams across time zones, and can experience high turnover. AI tools, on the other hand, can run without breaks, handle thousands of interactions simultaneously, and don’t request raises or vacation time.
Support interactions often follow set patterns. Customers ask about order status, password resets, shipping delays, refund policies, and technical glitches. These are issues that AI systems, once properly trained, can answer consistently and instantly.
When used carefully, AI in support roles can improve consistency, eliminate wait times, and allow human agents to focus on more complex or emotionally charged cases. It can also give smaller companies a way to offer 24/7 help without the overhead of staffing global teams.
The Human Cost
Still, the reduction in staff at Dukaan raises important concerns. Jobs in customer support are often entry-level roles. For many, they are a way into the tech industry or a first step toward more advanced positions. Replacing these roles with machines cuts off that entry point for future workers.
More importantly, not every interaction can—or should—be handled by a machine. Angry customers, emotionally distressed individuals, or those struggling to communicate clearly often need a human on the other end of the line. Machines are improving, but the ability to read between the lines, show patience, or offer genuine comfort still belongs to people.
Some customers have already begun to complain that AI responses feel too mechanical or miss the point entirely. This can erode trust, especially when people feel their concerns aren’t being truly heard or understood.
How Should Companies Move Forward?
The Dukaan case invites deeper reflection. Should companies be allowed to automate as much as they like, regardless of the social consequences? Or do they have some responsibility to retrain, reassign, or at least fairly compensate those who are replaced?
There are no easy answers. Technology has always displaced certain types of labor. But the speed and scale of current changes make it different from past shifts.
Transparency matters. So does tone. A company that must reduce its staff for efficiency reasons can still treat its employees with dignity. It can offer severance, career support, or even share its long-term vision in a way that shows respect for the people who helped it grow.
Announcements that focus solely on numbers, charts, and technical performance risk alienating not only staff but also customers who care about how businesses treat people.
A Year Later
About a year after the shift, reports indicate that Dukaan’s AI-powered support system continues to function effectively. Resolution times remain low, and support coverage has stayed consistent.
However, the backlash to the original announcement remains a cautionary tale. The event has become a talking point in discussions about automation—not because of technical failure, but because of how the change was handled publicly.
Meanwhile, other companies appear to be taking a more balanced approach. Some are experimenting with blended teams, where AI handles initial queries and human agents step in for more nuanced issues. Others are using automation to support their workers, not replace them outright.
Final Thoughts
The decision to cut 90% of a support team in favor of AI may look smart on a balance sheet. But the way such decisions are framed—and the consequences they carry—can shape public opinion for years.
Technology can improve how we operate, but how we choose to implement it defines the kind of workplace, and society, we build. The case of Dukaan is not just about one company’s chatbot. It’s a window into the choices leaders make when speed and scale meet people’s lives.
Businesses that value trust, loyalty, and long-term relationships will need to think carefully not only about what they automate, but how they communicate those choices to the world.