All roads lead to BPOs
Services, Economy
In an economy where services account for over half of GDP and employ hundreds of millions of people, AI's biggest opportunity isn't replacing software—it's replacing workflows.
India's Competitive Advantage Was Never Software. It Was Services.
For the last two decades, Indian startups have largely borrowed Silicon Valley's playbook: build software, charge subscriptions, and scale through SaaS. While companies like Freshworks, Zoho, Postman, and BrowserStack proved that India could build globally competitive software businesses, they are the exception rather than the rule. India's economy was never built on software exports alone, it was built on services.
Services contribute over 54% of India's GDP, employ nearly one-third of the country's workforce, and account for more than 40% of total exports. The IT and Business Process Management (BPM) industry alone generates over $280 billion in annual revenue, employing more than 5.5 million people directly. Outside technology, healthcare, banking, insurance, logistics, accounting, legal services, education, and retail collectively represent trillions of dollars in economic activity, yet much of this work remains manual, fragmented, and dependent on human coordination.
This is precisely where AI changes the equation.
Unlike the United States, where AI is often positioned as a productivity tool for highly paid knowledge workers, India's opportunity is fundamentally operational. Our largest industries are constrained not by a lack of software but by shortages of skilled professionals, inconsistent service quality, paperwork-heavy workflows, and low productivity per employee. AI doesn't merely automate tasks, it allows expertise to scale. A single doctor can supervise more patients. A loan officer can process significantly more applications. An insurance underwriter can review claims faster. A legal associate can draft documents in minutes instead of hours. India's competitive advantage has never been writing software for the world; it has been delivering services to it. AI simply makes those services infinitely more scalable.
The Next Generation of Indian Startups Will Sell Outcomes, Not Software
For decades, Indian IT giants like TCS, Infosys, Wipro, HCLTech, and Tech Mahindra built multi-billion-dollar businesses by taking ownership of business processes rather than selling products. They weren't merely software vendors—they became extensions of their clients' operations. AI dramatically changes the economics of that model.
Instead of hiring 500 analysts to process invoices, an AI-native company may need 20 experts supervising autonomous systems. Instead of licensing an AI copilot to hospitals, startups can charge per patient successfully triaged. Rather than selling an OCR tool to banks, they can guarantee that a loan application is verified within five minutes. The software itself becomes invisible; customers buy completed work.
This represents a shift from Software-as-a-Service (SaaS) to Service-as-Software.
We're already seeing early signs of this transition. AI is reducing medical documentation through ambient clinical assistants, automating insurance claims processing, accelerating legal contract review, improving logistics planning, and transforming customer support. The companies that capture the most value will not necessarily have the best large language model. Foundation models are rapidly becoming commodities. Open-source models continue to improve, inference costs continue to decline, and APIs are increasingly interchangeable.
The real moat lies elsewhere: proprietary workflows, operational integration, access to domain-specific data, regulatory knowledge, and the trust required to execute business-critical services. AI becomes the engine, but the business is defined by the outcome it consistently delivers.
India's Next AI Giants Will Look Like Service Companies
Much of India's AI conversation remains fixated on foundation models—how many parameters a model has, whether it's multilingual, or whether India can produce its own equivalent of OpenAI or Anthropic. While sovereign AI capabilities matter strategically, they are unlikely to define where most enterprise value is created.
Historically, the largest technology companies emerged by solving economic bottlenecks rather than competing on underlying infrastructure. Amazon built logistics before cloud computing. Stripe abstracted payments, not banking. Nvidia's value today comes not merely from chips but from enabling an entirely new computing paradigm.
India's bottlenecks are different.
Healthcare faces a doctor-to-population ratio below WHO recommendations, while millions still lack access to consistent primary care. Insurance penetration remains low, yet claim processing is often slow and administrative. India's logistics costs are estimated at 13–14% of GDP, significantly higher than developed economies, largely because of operational inefficiencies. The legal system suffers from enormous case backlogs, financial institutions continue to rely on manual compliance processes, and millions of small businesses still operate without sophisticated operational software.
These are not software problems. They are service delivery problems.
The companies that define India's AI decade will likely resemble traditional service businesses from the outside. They may employ doctors, lawyers, accountants, logistics operators, or financial experts. But internally, their operations will increasingly be orchestrated by AI systems capable of performing the majority of routine cognitive work while humans supervise, validate, and handle edge cases.
This is why India's AI opportunity is fundamentally different from Silicon Valley's. America built software because software was its comparative advantage. India built services because services were ours. AI is now collapsing the distinction between the two.
The next generation of Indian unicorns won't simply sell software to businesses. They will become the businesses themselves, using AI to deliver healthcare instead of hospital software, underwriting instead of underwriting tools, legal work instead of legal software, financial operations instead of fintech infrastructure.
