For years, Silicon Valley focused largely on software. Apps, platforms, and digital marketplaces captured investor attention—and billions in venture capital. But a quieter shift is underway. Some of the world’s largest and oldest industries—construction, manufacturing, and logistics—are increasingly being viewed as the next frontier for technology investment.
These sectors may not feel like “tech,” but together they represent trillions of dollars in economic activity. As artificial intelligence, robotics, and automation become capable of operating in complex, real-world environments, the opportunity to modernize these legacy industries is enormous.
To better understand why under-digitized sectors are emerging as trillion-dollar technology opportunities, I spoke with Nikhil Choudhary, Managing Partner at Nirman Ventures, a Silicon Valley-based investor and operator focused on technologies modernizing the physical economy. He’s spent over two decades building and scaling businesses across construction and real-asset-driven industries, before turning his attention to early-stage investing in robotics, autonomy, and applied AI systems deployed in real-world environments.
Choudhary’s work centers on under-digitized sectors where productivity, safety, and labor constraints present both structural challenges and long-term technology opportunities. His perspective is shaped by hands-on operating experience, with an emphasis on solutions that perform reliably in uncontrolled settings and deliver measurable outcomes on job sites, in warehouses, and across industrial operations.
Rieva Lesonsky: Why are legacy industries, like construction and manufacturing, suddenly the next trillion-dollar tech opportunity?
Nikhil Choudhary: What’s changed is that the world’s biggest industries are finally becoming computable. Construction alone is a ~$11T market today and headed toward ~$16T by 2030, according to Deloitte. If you add up the size of the construction, manufacturing, and logistics markets, you’re talking about a massive amount of money. Even a small improvement of one or two percent could mean hundreds of billions of dollars.
For a long time, it was hard to make big changes in these industries because a lot of the work was done by people, and profits were small. There’s also a significant shortage of workers in these traditional industries, and at the same time, a lot of money is being invested in modernizing them, partly because of changes in infrastructure, energy, and government policies.
What’s different now is that artificial intelligence is good enough to work in the real world, not just in perfect test situations, but in the messy, physical environments where these industries actually operate. This means AI can really start making a difference in these industries, and that’s a big deal.
Lesonsky: How did your experience as an operator shape your approach to investing in deep tech and robotics?
Choudhary: Being an operator changes what you care about. Once you’ve had to deliver outcomes, manage cash, and deal with real customers, you start asking practical questions. I look at new technologies, like robotics, in a practical way. I don’t just think about how advanced they are, but about whether they can work in real-world situations, where things can get messy, and customers can be slow to adapt.
And let’s be honest, when downtime costs money, customers get nervous. Being an operator also gives you a sense of timing. You realize that developing new technologies takes time, and that’s okay. What matters is that the company is making progress, reducing risks, and getting closer to its goals with each step. It’s not just about being fast; it’s about being real.
Lesonsky: What is “physical AI,” and how does it differ from software-only AI?
Choudhary: Physical AI refers to artificial intelligence systems integrated with hardware that interact with the physical environment, providing tangible outputs and requiring accountability for real-world actions, unlike software-only AI, which operates in virtual environments. When software AI makes a mistake, you get bad output. When physical AI gets it wrong, something breaks, stops, or becomes unsafe.
Lesonsky: Why have many past attempts to digitize or automate legacy industries failed?
Choudhary: It’s tough for physical industries to change. They have small profits, it takes a long time to make sales, and they have to worry about safety and legal issues. Also, their operations cannot afford to stop working, which makes it hard to try new things without risking everything. This makes it slow and difficult to make changes in these industries. Things are still tough, but the situation has changed. Now, new technologies and major infrastructure investments are driving progress. At the same time, it is getting cheaper to put new systems in place, and we can see the benefits of automation more clearly. That’s what makes this time different. We’re not just talking about spending money; we’re talking about getting results. And with workers in short supply, companies are looking for new ways to get things done. This is why automation is becoming more important. It’s not just about cutting costs; it’s about being able to do more with less. And that’s a big change from how things used to be.
Lesonsky: How do operator-led investment funds differ from traditional Silicon Valley venture funds in their approach?
Choudhary: Traditional Silicon Valley venture funds and operator-led funds play different but complementary roles. Platform funds are incredibly good at pattern recognition, scaling playbooks, and building companies once a market is already well understood. There’s a reason they’ve been so successful.
When new technology enters the market, our type of operator-led fund usually gets involved early on. This is especially true in industries such as construction, manufacturing, and logistics. The challenge isn’t just about the technology itself, but also about understanding how things are done, where problems arise, and how people adopt new ways of doing things. As technology increasingly changes more industries, the knowledge and experience that come from being an operator become more important.
Lesonsky: What practical lessons can founders learn when applying AI and robotics to traditional industries?
Choudhary: Old-fashioned businesses can’t keep up with the fast pace of the tech world. They just can’t afford to rush things. Sales cycles are long because decisions carry real liability, safety risk, and regulatory exposure. One mistake can mean downtime, punitive damages, or injury.
Time is a big deal: We’re all busy, mistakes can be costly, and most agreements are set up to address problems that might arise. And finally, these industries employ large, highly skilled blue-collar workforces that are sensitive to economic cycles. Founders who succeed respect this reality and build tools that support the workforce, fit existing workflows, and earn trust gradually rather than trying to force change all at once.
Lesonsky: Where do you see the next generation of billion-dollar deep tech companies emerging globally?
Choudhary: Where billion-dollar companies emerge is usually a function of two things: How much capital is being deployed and whether the market is ready to absorb new technology. Comparing this through numbers: In 2025, India’s deep tech sector raised ~$1.65B, China raised ~$81B, while the US raised ~$147B over the same period. Those gaps matter because deep tech needs patient capital, not just talent, to survive long deployment cycles.
I expect the U.S. to remain the primary home for the next wave of large deep tech companies for the near future, largely because of its capital depth and appetite for risk. Followed closely by China due to various other factors that free markets can’t compete with. That said, we’re also starting to see meaningful momentum build in Japan and India. As capital, talent, and national priorities align in those markets, their presence in deep tech will grow quickly.
Lesonsky: How do you evaluate founders and startups in industries that are often misunderstood by investors?
Choudhary: The best teams can explain the problem in plain language, describe where things break in real operations, and show that they’ve earned trust from customers who are typically skeptical of new technology.
I take a close look at whether the people who started the company have actually experienced the problem they’re trying to solve, either by working in the field themselves or by spending a lot of time with the people they’re trying to help. The really great founders are completely focused on understanding what their customers need. And in industries that have been around for a while, it’s not just about having new ideas—it’s also about being credible and trustworthy.
Lesonsky: What role do global capital, partnerships, and ecosystem support play in scaling physical-world technologies?
Choudhary: For physical-world technology to scale, you need the right partners and ecosystems to get into the field, test in real conditions, and earn customer trust.
From an operator’s perspective, global capital matters because it unlocks access to manufacturing, supply chains, pilot sites, and customers who are willing to take early risk. The teams that scale fast are the ones that build those relationships early and use the ecosystem to move from pilot to production, not just from demo to demo.
Lesonsky: Can you share a story of a portfolio company that successfully implemented robotics or AI in a legacy sector?
Choudhary: Yes, let me share two:
Citian, a Washington, D.C.-based company where we were the first institutional investors, is integrating AI and real-world data to enable smarter planning for transportation planners and engineers. Citian’s solutions are currently used by several dozen cities, counties, states, and transportation departments. Citian is now at a growth stage with millions in revenue.
Noble Machines, a Silicon Valley-based robotics company, manufactures rugged humanoid robots for traditional industries. They have deployed and shipped humanoid robots to customers in and outside the United States.
Rieva Lesonsky is the founder of Small Business Currents, a content company focusing on small businesses and entrepreneurship. You can find her on Twitter @Rieva, Bluesky @Rieva.bsky.social, and LinkedIn. Or email her at Rieva@SmallBusinessCurrents.com.
Photo courtesy Noble Machines
