The Surfboard Factory

Vishal Sikka just raised $32 million for Hang Ten Systems, an enterprise AI implementation company. He is building it from Palo Alto, not Bangalore.Let that sink in.This is the same Vishal Sikka who was CEO of Infosys—the company that proved to the world that enterprise technology could be built, delivered, and scaled from India. The man who sat in Bangalore boardrooms and convinced Fortune 500 CEOs to move billions of dollars of IT work to the subcontinent. The architect of the Indian IT services model.And now, when he looks at the biggest technology shift of our lifetimes—AI—he plants his flag in California.That is not a coincidence. It is a first-principles signal about what AI implementation actually is. And it is the exact signal that led us to build Niyamic from India.Let me explain.

The Implementation Paradox

For thirty years, the Indian IT services model worked on a simple formula: American judgment + Indian labor = enterprise value.The strategy, architecture, and client relationships lived in the US. The coding, testing, maintenance, and scaling lived in India. The arbitrage was labor cost. The value was scale.That model built a trillion-dollar industry. It built Infosys, TCS, Wipro, and HCL. It built the modern Indian middle class.But AI has killed the premise.When Vishal Sikka says Hang Ten will help enterprises "build in days what used to take years," he is not talking about moving coding work to a cheaper geography. He is talking about something far more radical: the marginal cost of software production is collapsing toward zero. Agentic code generation, reusable AI skills libraries, and frontier models are doing to custom software development what containerization did to shipping—they are unbundling the work from the workers.If AI can write the code, the arbitrage is no longer "cheaper humans." The arbitrage is better judgment.And judgment lives in the boardroom, not the back office.

Why Palo Alto? The Geography of Trust

Look at Hang Ten's first customers: Fresenius. Siemens Energy. These are not companies looking for cheaper Java developers. These are industrial giants trying to figure out how AI changes the fundamental economics of dialysis, wind turbines, and energy infrastructure.This is not an outsourcing conversation. This is a transformation conversation.Sikka's LinkedIn post reveals the truth in one line: "I have seen, firsthand, the dramatic things AI delivers for the people and teams who somehow just know what to do with it—I have watched them, and myself, reach in minutes what could take teams years of toil. And I have seen the far greater number who get none of it, and who often end up causing harm instead."That gap is not a labor gap. It is a judgment gap.It is the gap between knowing which enterprise processes to automate, which legacy systems to replace, which data to trust, and which hallucinations to fear. It is the gap between playing with ChatGPT and rewiring a $50 billion healthcare supply chain. It is the gap between experimentation and transformation.Closing that gap requires something that cannot be Zoomed in from Bangalore: trust at the highest levels of the enterprise. It requires sitting across from the CEO of Fresenius and saying, "We are going to rewire how your company runs, and we are going to do it with technology that did not exist eighteen months ago."That conversation happens in Palo Alto, Munich, and Houston. It does not happen over a TCS-style "delivery center" model.Mayfield and Aramco Ventures did not write a $32 million seed check to a company that would win on labor arbitrage. They wrote it to a company that would win on transformation trust. That is a Palo Alto game.

So Why Is Niyamic Building from India?

If judgment lives in Palo Alto, why are we building Niyamic from India?Because we are playing a different—but equally first-principles—game.Hang Ten is solving the judgment problem: "What should we build, and how do we convince the board to let us build it?"Niyamic is solving the architecture problem: "Now that we know what to build, how do we actually make it work at scale, reliably, in production, across global enterprises?"Here is the counterintuitive truth: AI implementation is splitting into two distinct disciplines.The first discipline is transformation consulting—the high-trust, high-judgment, board-level work of figuring out where AI creates value. This is what Hang Ten does. It is McKinsey meets Accenture Strategy meets frontier AI research. It requires proximity to power, decades of enterprise relationships, and the rare ability to translate AI hype into CFO-friendly ROI. This is why Sikka is in Palo Alto.The second discipline is AI-native engineering at scale—the work of wiring models into ERP systems, building agentic workflows that do not hallucinate, fine-tuning on proprietary data, and running the MLOps infrastructure that keeps enterprise AI from becoming enterprise AI disaster. This is what Niyamic does. And this is where India has a structural advantage that has nothing to do with cost.

India's New Arbitrage: AI-Native Talent Density

The old Indian IT advantage was: "We have more engineers who cost less."The new Indian AI advantage is: "We have the highest density of AI-native engineers on the planet."Not "engineers who learned AI." Engineers who were born into it. Engineers who do not remember a world before transformers. Engineers who grew up on Kaggle, Hugging Face, and GitHub Copilot. Engineers who treat agentic code generation as a primitive, not a novelty.India produces more than 1.5 million engineering graduates every year. A staggering percentage of the world's Kaggle Grandmasters, arXiv AI paper authors, and open-source ML contributors are Indian. The Indian Institutes of Technology and the Indian Institutes of Information Technology are becoming the world's largest factories of AI-native talent.But here is the critical distinction: this talent is not cheaper. It is different.An AI-native engineer in Bangalore does not cost 1/5th of a Silicon Valley engineer anymore. The gap has narrowed to perhaps 1/2 or 2/3. But the type of work they can do is fundamentally different from what a traditional offshore team could do.Traditional offshore teams executed specifications. AI-native engineers in India today are co-architecting agentic systems. They are building reusable AI skills libraries. They are designing retrieval-augmented generation pipelines for Fortune 500 data. They are the people who take the "judgment" from Palo Alto and turn it into running software.This is not cost arbitrage. This is capability arbitrage.

The Surfboard Factory

Sikka's surfing metaphor is perfect. Let me extend it.Palo Alto is where the waves break. It is where you learn to read the ocean, time the swell, and convince the boat captain to let you jump in. It is where Hang Ten teaches enterprises to "hang ten"—to master the wave so completely that they can walk to the front of the board and hang their ten toes off the front.But someone has to build the surfboards.Someone has to shape the fiberglass, design the fins, and test the boards in a thousand different conditions so that when the surfer paddles out, the equipment does not fail.That is Niyamic. We are building the surfboard factory.We are not trying to sit in the boardroom with the CEO of Fresenius. We are building the AI-native engineering infrastructure that makes Hang Ten's vision—and the vision of every enterprise AI transformation—actually work in production.We are building the reusable skills libraries. The agentic workflow engines. The enterprise-grade RAG systems. The fine-tuning pipelines. The MLOps infrastructure that keeps AI from becoming a $20 million proof-of-concept that never ships.And we are building it from India because that is where the surfboard engineers live.

The Generational Inversion

There is a generational story here that most people are missing.Generation 1 (1990s–2010s): America had the judgment. India had the labor. Value flowed from US strategy to Indian execution. The model was: "Tell us what to build, and we will build it cheaper."Generation 2 (2020s–now): America still has the judgment—but now it also has the implementation layer for high-trust transformation. India is watching from the shore as the biggest AI services companies are built in Palo Alto, not Bangalore. The model is: "We will figure out what to build, and we will build it here."Generation 3 (what Niyamic is building): America will always have the judgment. But India is building the AI-native capability layer that makes global implementation possible. The model is: "We will co-architect what to build, and we will build the infrastructure that makes it scale anywhere."This is the inversion. The last generation of Indian IT services was about labor cost. The next generation of Indian AI services is about talent density. We are not the back office anymore. We are the engineering core of the AI implementation stack.

Why This Matters

Vishal Sikka launching from Palo Alto is not a rejection of India. It is a clarification of what AI implementation actually is.It tells us that the first layer of enterprise AI—the judgment layer, the trust layer, the transformation layer—is a proximity game. It requires being in the room where the wave breaks.But it also tells us that the second layer—the architecture layer, the engineering layer, the "make it actually work" layer—is a capability game. And that game is increasingly being won by the geographies with the highest density of AI-native talent.That is why Hang Ten is in Palo Alto. And that is why Niyamic is in India.We are not trying to be the next Infosys. Infosys was built on a model that AI has made obsolete. We are trying to be the first Niyamic—a company built on the first-principles truth that AI implementation is not about cheaper labor. It is about better architecture.And the best architecture is being built by the engineers who have never known a world without AI.Most of them live in India.If you are building enterprise AI and need a team that can turn judgment into architecture, we should talk.Niyamic — AI-native implementation, built from India for the world.