Deep within a bustling Indian startup office, a team of dedicated developers is meticulously refining artificial intelligence chatbots, crafting them to communicate with an uncanny resemblance to human interaction.
This ambitious company, LimeChat, has a bold vision: to render most customer service roles redundant. They assert that their advanced generative AI agents can drastically reduce the human workforce required to manage up to 10,000 monthly customer inquiries by a staggering 80%.
Nikhil Gupta, the 28-year-old co-founder of LimeChat, shared a striking claim: ‘Once you onboard a LimeChat agent, your hiring needs for that role effectively cease.’
For years, India’s cost-effective labor and widespread English proficiency cemented its status as the global hub for back-office operations, often impacting employment in other regions. Today, a new wave of change is sweeping through this sector. A recent Reuters investigation reveals that AI-powered systems are now rapidly absorbing roles traditionally performed by university graduates in areas like technical support, customer care, and data management, forcing a widespread re-evaluation and adaptation across the industry.
This accelerating trend is fueling a boom for AI startups, which offer companies the ability to drastically cut staffing expenses and scale operations, even as a significant number of consumers continue to express a preference for human interaction.
This detailed report on the transformative shifts within India’s $283 billion IT sector draws from extensive interviews with 30 individuals, including leading industry executives, recruitment specialists, affected workers, and both current and former government officials. The analysis also includes firsthand tests of advanced voice and text chatbots, demonstrating their capacity for increasingly complex and human-like customer engagements.
Instead of slowing down in the face of AI’s threat to routine jobs, India is aggressively pushing forward, betting that a rapid embrace of the technology will generate sufficient new opportunities to re-employ those displaced. The outcome of this strategic gamble holds immense significance globally, serving as a critical test case for whether AI-driven disruption can truly uplift a developing economy or ultimately become a cautionary example.
Experts at Grand View Research project that the global conversational AI market is expanding at a remarkable 24% annually, anticipated to hit a valuation of $41 billion by the year 2030.
India, where the IT sector contributes a substantial 7.5% to its GDP, is fully committing to this technological transformation. Prime Minister Narendra Modi, in a speech given in February, articulated his belief that ‘technology doesn’t eliminate work; it merely alters its nature, leading to the creation of novel job categories.’
However, not all observers share Modi’s optimistic outlook regarding India’s readiness. Santosh Mehrotra, a former Indian official and a visiting professor at the University of Bath’s Centre for Development Studies, expressed concern, criticizing the government’s perceived lack of urgency in evaluating AI’s potential impact on the nation’s young workforce. He frankly stated, ‘There’s no clear strategy.’
The business process management sector in India currently employs 1.65 million individuals across call centers, payroll, and data handling. According to Neeti Sharma, CEO of the staffing firm TeamLease Digital, hiring in this segment has significantly decreased due to widespread automation and digitalization, even amidst a growing demand for specialized roles like AI coordinators and process analysts.
TeamLease Digital figures illustrate this stark change: net employment within this segment, which contributes one-fifth of India’s total IT output, saw an increase of less than 17,000 workers in each of the last two years. This is a dramatic drop from increases of 130,000 in 2022-2023 and 177,000 in 2021-2022.
Interviews conducted by Reuters with three current and five former customer service employees reveal a growing sense of job insecurity, directly linked to the increasing integration of AI. These tools range from systems that suggest responses to fully autonomous bots handling almost all routine inquiries. Megha S., a 32-year-old who previously earned $10,000 annually at a Bengaluru-based software solutions provider, recounted being laid off just last month, right before India’s festive season. Her company was transitioning to AI tools for quality review of sales calls, leading to her displacement.
Megha, who requested anonymity regarding her full name and former employer, shared her personal struggle: ‘I was informed that I was the very first person to be replaced by AI.’ The weight of this news is evident as she added, ‘I haven’t even told my parents yet.’
Sumita Dawra, a former labour ministry secretary who previously headed an Indian government task force examining AI’s impact on the workforce until her retirement in March, acknowledged that while AI brings significant productivity advantages and will create new jobs, she believes India should explore more robust social security measures, such as unemployment benefits, to support workers displaced during this transitional period.
In contrast, a senior Indian official, who spoke to Reuters, indicated that the government maintains a stance that AI’s long-term effect on overall employment will be minimal. Requests for comments to India’s IT and labor ministries, as well as the Prime Minister’s office, went unanswered.
Beyond the rise of AI, several other factors cast uncertainty over India’s IT sector. These include potential U.S. tariffs, a legislative proposal from a U.S. lawmaker for a 25% tax on companies utilizing foreign outsourcing, and a significant $100,000 fee imposed by President Trump on new H-1B visas, a common pathway for tech firms to sponsor Indian talent.
In September, investment bank Jefferies forecasted a substantial financial impact on India’s call centers, predicting a 50% revenue decrease due to AI adoption over the next five years. Other back-office functions are also expected to see a decline of approximately 35%.
This outlook suggests imminent job losses across India, a country that currently commands a dominant 52% share of the global outsourcing market.
Pramod Bhasin, who pioneered India’s first call center in the 1990s with just 18 employees for GE Capital – where makeshift saris often served as workstation dividers – remarked that ‘the most significant impact will be felt by young college graduates entering the workforce.’
Looking further ahead, Bhasin, who later founded the IT services giant Genpact, suggests that India has the potential to evolve from being the world’s ‘back office’ to becoming its ‘AI factory.’ This transformation could be realized by strategically capitalizing on the increasing demand for AI engineers and expertise in automation deployment.
LimeChat is one startup thriving on this demand, as observed by Reuters during an August visit. Co-founder Gupta stated that his team of developers and engineers has successfully automated 5,000 jobs throughout India. Currently, their bots manage 70% of customer complaints for their clients, with an ambitious goal of reaching 90-95% within the next year.
Gupta elaborated on the cost-efficiency, explaining, ‘If you’re paying us 100,000 rupees per month, you’re effectively automating the work of at least 15 agents.’ He noted that this monthly cost, approximately $1,130, is roughly equivalent to the salary of just three human customer-care employees.
Regulatory disclosures indicate a remarkable surge in LimeChat’s sales, climbing from $79,000 just two years prior to an impressive $1.5 million in 2024. Last year, the company forged a partnership to integrate Microsoft’s Azure language models and algorithms, leading to the launch of an innovative new e-commerce chatbot.
Among Gupta’s growing client list is Kapiva, an Indian firm specializing in Ayurvedic products, which has successfully implemented a LimeChat bot to manage customer interactions via WhatsApp.
During a demonstration, typing the prompt, ‘What kind of diet should I have to reduce weight?’ instantly produced an AI-generated meal plan. A subsequent query, posed in both English and Hindi, asking about the distinction between a slimming juice and another product, was also expertly answered. The chatbot concluded by cheerfully offering links to relevant Kapiva products, complete with a smiling emoji.
LimeChat faces competition from industry giants like Reliance, the conglomerate led by Mukesh Ambani, which acquired the Indian startup Haptik in 2019.
Haptik proudly advertises its ‘AI agents’ as capable of delivering ‘human-like customer experiences’ at a cost of $120, claiming they can reduce support expenses by 30%. Financial disclosures reveal a phenomenal growth in revenue, surging from under $1 million in 2020 to nearly $18 million last year.
In a September webinar, Haptik provocatively promoted its services by asking participants, ‘Imagine having a tireless, full-time employee who costs merely 10,000 rupees per month?’
Suji Ravi, Haptik’s product manager, highlighted the significant industry transformation during the webinar, stating, ‘We are witnessing a massive shift. Companies are increasingly choosing to invest in AI agents rather than hiring more human staff.’
For Mamaearth, an Indian personal-care brand and a client of LimeChat, the primary appeal of AI chatbots lies in their unparalleled scalability, explained Vipul Maheshwari, Head of Product and Analytics at Mamaearth’s parent company, Honasa Consumer.
Maheshwari emphasized the critical role of customer support, stating, ‘Delivering excellent customer service is crucial for our success. However, the question remains: can we expand our human customer support team indefinitely? The answer is a clear no.’
The Mamaearth chatbot, according to Maheshwari, can perform more than just basic tasks like order tracking. It can guide users with complex inquiries, such as recommending suitable products during pregnancy, and even skillfully manage interactions with upset customers.
The dual nature of AI – its immense promise and potential pitfalls – is starkly illustrated at The Media Ant, a Bengaluru-based advertising agency. Founder Samir Chaudhary revealed that the firm reduced its workforce by 40% to approximately 100 employees over the past year, and also gave up office space in another building to cut rental costs.
Chaudhary detailed the changes, explaining that the company replaced 15 salespeople with AI bots designed to identify leads and dispatch emails to potential clients. Furthermore, a six-member human call center was superseded by an AI voice agent named Neha, who communicates in remarkably fluent, Indian-accented English.
During a test, when a Reuters reporter inquired about YouTube advertising, Neha efficiently gathered information on budget and target markets, recorded the requirements, and concluded the conversation with a cheerful sign-off: ‘I will email you the details… have a great day.’
Chaudhary playfully noted, ‘If you were to ask her out for coffee, she’d simply brush it off with a laugh.’
However, the enthusiastic adoption of AI doesn’t always proceed without complications for businesses. Sweden’s fintech company, Klarna, offers a case in point. While chatbots enabled the firm to eliminate thousands of jobs last year, its CEO disclosed in September that Klarna is now actively ‘trying to course correct,’ shifting its focus towards leveraging AI to enhance products instead of solely cutting costs.
Indeed, chatbots are not without their limitations. While LimeChat’s bots successfully addressed most general e-commerce inquiries during Reuters’ testing, certain complex questions proved challenging for them.
For example, when Knya’s bot, a LimeChat client, was asked to substantiate its assertion that a million medical professionals rely on its products, like stethoscopes, it responded with, ‘I am sorry, I don’t have enough information to answer your question.’
Despite their advancements, customer surveys consistently reveal that many users still harbor a strong dislike for chatbots.
A survey conducted by EY in August 2024, involving 1,000 Indian consumers, indicated that 62% of respondents made purchasing decisions influenced by AI recommendations, significantly higher than the global average of 30%. Nevertheless, EY observed that ‘the longing for human connection persists strongly,’ with 78% of consumers favoring online platforms that offer human assistance.
Despite these findings, LimeChat’s Gupta argues that properly trained AI agents possess the capability to resolve inquiries more swiftly than humans. He noted that while many standard bots escalate interactions to human agents when faced with irate customers, ‘you only need a very small team to manage these negative experiences.’
The tech boom in India during the 1990s and 2000s spurred a significant rural-to-urban migration. Cities such as Bengaluru rapidly transformed into major outsourcing centers, as Indian companies like Tata Consultancy Services, Infosys, and Wipro evolved into global industry powerhouses.
This expansion had a cascading effect, reaching neighborhoods like Ameerpet in Hyderabad, where classrooms are consistently filled with university graduates eager to acquire IT skills and certifications for tech roles.
Historically, Ameerpet’s training centers primarily offered courses in software like Microsoft Office and programming languages such as Java. However, a Reuters visit in April revealed a significant shift, with these centers now increasingly concentrating on specialized AI training.
Outside one such institution, Quality Thought, a prominent banner strikingly displayed a robot gazing over a globe, prominently featuring the letters ‘AI.’
This particular center was offering a comprehensive nine-month course in AI data science and prompt engineering for approximately $1,360, a price point more than double that of a conventional web-development program.
Priyanka Kandulapati, a staff member, confirmed the industry trend, stating, ‘Recruiters are actively seeking candidates with foundational AI skills.’ She added, ‘We are committed to further refining our course offerings to precisely meet this evolving demand.’
During a recent discussion with startup founders regarding the rapid pace of technological change, venture capitalist Vinod Khosla, a co-founder of Sun Microsystems, presented a sobering perspective on India’s future.
He candidly predicted, ‘Virtually all IT services will be automated or replaced within the next five years. This transition is going to be quite chaotic.’