Kanwal Rekhi, Silicon Valley veteran, co‑founder of Excelan and founding partner at Inventus Capital Partners, has spent over five decades building, backing and mentoring entrepreneurs. He is also the founder of The IndUS Entrepreneurs (TiE), a global network of Indian entrepreneurs. From being laid off from his first three jobs to becoming the first Indian‑American founder to take a venture‑backed company public—when Excelan, an early networking pioneer, listed in 1987—Rekhi has witnessed Silicon Valley’s evolution firsthand.
He has since mentored more than 10,000 founders, invested in 200-plus early‑stage companies, and backed ventures such as Poshmark, redBus, Exodus Communications, Sierra Atlantic and Netmagic. In a conversation with Forbes India at TiE’s Palo Alto office, the 80‑year‑old investor questions today’s AI‑led exuberance, and explains why founders must know when to hand over their companies to professional managers. Edited excerpts:
Early‑stage investors like me invest in about 10 companies, of which six or seven fail, and maybe two run profitably. I need exits to cover all my losses; that’s why exits matter. You have early‑stage investors, mid‑stage investors, and late‑stage investors, each with a different level of risk. Early‑stage investors need 50–100x returns; mid‑stage investors aim for 10–15x and late‑stage investors are happy with 3–5x, with very low risk.
If I don’t make 50x or 100x return as an early investor, I’m in the wrong business. The ecosystem in Silicon Valley supports this. The culture is deeply embedded. Everybody understands the process.
I’m always early‑stage, with highest risk. That’s been my life—it’s hard to change. My value is not just money; it’s finding entrepreneurs, finding markets, mentoring them. That gives me a different kind of joy.
Investors are putting money into AI, because it is very hot right now. There’s a sense that AI is hyped up right now. It’s a circular business: OpenAI is making a deal with Oracle, making deals with Nvidia. Nvidia is making deals with AMD. And investors who are making money there are investing it back into OpenAI. It just keeps closing the loop. As soon as the hype hits a wall, the air goes out of the system overnight.
There is no return right now. AI as a business does not yet have a profitable business model. They’re chasing scale. What do they sell to the consumer, $10 a month? But it’s costing them $20 a month to produce, because power and compute are very expensive.
As long as money is available, it will work. But money has a way of not being available. When the tide comes in, all the boats go up. When the tide goes out, all the boats go down. Right now, the tide is in, so everything looks fine.
I’m from the old school. At some reasonable size—reasonably small size—your model should start to work. I don’t mind you raising money and growing, but there has to be a path where, if the tide goes out, I stop growing, but I still have profitable operations.
Silicon Valley has deep technology roots. Fundamental technology was developed here—semiconductors from the very start, databases, operating systems, chips. India is not developing fundamental technology. And, by the way, that’s not bad; that’s a smart thing to do. Why would I risk my capital to compete with American technology when the market is global?
Take India’s IT industry. It provides services around technology developed elsewhere and has built very large businesses—Wipro, Infosys, TCS. They implement, they deploy, they grow. The technology comes from America. They take people, they train them, they use technology, and they deliver solutions. That’s a very good business—profitable, growing nicely, and the world needs it.
But if I say I want to develop a database to compete with Oracle, I have to spend millions of dollars developing it, and while I’m developing it, I have nothing to sell. It’s a bottomless pit. I’m just putting money in. By the time the database is ready, whether I can compete with Oracle or not—that’s still unclear.
Having said that, at some level, India as a country needs to become self‑reliant, like China is becoming. China used to depend on us for technology; now China is developing its own. But that requires huge investment. And the problem with India is there is no domestic capital for this type of investment. Service businesses can fund themselves—but can they fund AI or deep tech? There are two problems: One, there is no local VC [venture capital] money; two, government policies actively discourage foreign capital from coming into India. There is no basis to develop advanced technology in India.
The way to think about it is unit economics. I should be able to produce a unit of whatever I sell for $1, and I should sell it for $1.5. That is unit economics, which is profitable.
The reason I’m going to lose money initially is because, as I’m selling more, I’m hiring more salespeople, I’m marketing more, I’m setting up new factories. I need financing—and that’s fine. But if my unit economics are positive, then at some size, the positive margins will cover my other expenses.
But when companies sell a service where the units are not profitable, what’s the point? You’re selling a $1 bill for 99 cents. You can sell all the dollar bills you want—people will buy them out. You need to have a very clear model of ‘at what size I will become profitable’, and that’s the size you shoot for. Then you figure out how much money you need to raise to reach that size.
But if you keep changing the goalposts—I’ll double, I’ll triple—you can never become profitable.
Exits happen for that reason alone. Early investors, early founders need to cash out the value they’ve created, and then hand off the business. They can stay invested with part of their wealth, but not all of it. Exits are not the end. They’re just another point along the journey—a way forward.
Entrepreneurs are restless. They don’t enjoy discipline. They disturb markets, break products, find new ways of doing things. A professional CEO’s job is different. Make sure trains run on time. Make sure inventory is shipped. Make sure invoices are raised, money is collected. It’s an entirely different mindset—very structured, very objective.
The exciting, disruptive work versus the disciplined, repetitive work—it’s nearly impossible for the same person to do both equally well. Can the person who achieved 10x growth suddenly be happy with 5-10 percent improvement in margins? Usually not. As a founder, you need to know when to bring in management.
I’m 80 years old and entrepreneurs still like working with me. I’m still able to add value. I do worry sometimes—one day my advice might become useless, based on old assumptions. It’s a changing world. Everything has to change—your models, your rules, your assumptions. At every step, one has to stay relevant.
First Published: Feb 09, 2026, 18:35
Subscribe Now(This story appears in the Feb 06, 2026 issue of Forbes India. To visit our Archives, Click here.)