Jagdeep Singh (Rhoda AI) — Building and...

1. Key Themes

The Four-Part Framework for Building Great Companies

Jagdeep has distilled decades of company-building into a repeatable checklist he applies before starting every venture. He insists all four elements must be present before committing.

"The first is you have to find a large unsolved problem... The second thing you need is you need to have a differentiated solution... The third key thing, obviously, goes without saying, is you really do need a world-class team... And the last thing that I look for in my companies... I look for early customer validation." [00:02:38]

"The answer I want to hear is, well, this is so compelling that we want to help you get this to market. What can we do to help you make this thing real? That's when you know you got something interesting." [00:04:23]

Contrarian Thinking as a Prerequisite for Startup Success

Jagdeep draws an explicit parallel between value investing and entrepreneurship — the insight being that consensus validation is actually a warning sign, not a green light.

"In investing, there's a notion of being contrarian, right? If you simply invest in stocks that everybody thinks are great stocks, if that's the conventional wisdom, then it's already priced in. And even if you're right, there's not a lot of value to be created. I think the same applies to entrepreneurship, right? To ideas. You want to pick an idea that's contrarian, which means that not everybody you talk to will agree that it's a good idea. In fact, if everybody says it's a great idea, it's probably already too late because it's already priced in, if you will." [00:10:43]

Build for Permanence, Not for Exit

Jagdeep explicitly rejects M&A as a founding thesis, framing it as a failure of ambition. The Latera/Sienna story is his cautionary tale — the company that sold for $550 million ended up generating billions for the acquirer.

"Every company I start now, I don't start it with the intention of M&A. I started with the intention of building a permanent company that can really make an impact in the industry... If you start a company with M&A as the end goal, then you're just not thinking big enough, in my view, to really make a big impact." [00:14:30]

Deep Tech Requires Long Time Horizons

Both Invenera (nearly 10 years) and QuantumScape (14 years) reflect Jagdeep's belief that material science and hard technology companies cannot be rushed. He sees this as a feature, not a bug.

"The technologies in both companies were deep tech, kind of difficult technologies. They both involved material science and, you know, semiconductors in one case and battery fabrication in the other... My preference is to build companies for the long run. So I would like to do companies for, you know, 10 plus years at a stretch." [00:20:04]

Robotics as the Last Major Unsolved AI Frontier

Jagdeep's thesis for Rhoda is that every other major AI application domain already has a dominant incumbent — making robotics the last large white space.

"In language models, in image generation models, video generation models, there already existed 800-pound gorillas in each of those spaces... The one area where AI had not yet made an impact, where there was no 800-pound gorilla, was in robotics... Nobody had figured out how to take robots from the laboratory, where they can do kind of cool things, into the real world." [00:21:57]

The Real World Distribution Shift Problem in Robotics

Jagdeep articulates a specific, technical reason why AI that works in every other domain fails in robotics — a diagnosis that frames his entire company's mission.

"In the laboratory setting, you can basically control the setting that the robot is operating in and allow the setting to match the data set that you train the model on very closely. In the real world, the actual data set that you see diverges from the training data set. And that divergence is enough to make the models fail." [00:22:55]

The CEO as Chief Risk Officer

Jagdeep reframes the founding CEO role entirely — not as visionary or salesperson, but as the person responsible for systematic risk reduction.

"Your job as CEO or co-founder is to be the chief risk officer of the company, which means you've got to be thinking about how many ways can this go wrong. And then for every one of those different ways it can go wrong, you need someone who's worrying about it more than you are... Usually it's not the risks you worry about that come back to bite you. It's the risk that you aren't worrying about that actually cause a problem." [00:32:17]

Investor Selection as a Four-Dimensional Decision

Jagdeep treats investor selection with the same rigor as hiring, evaluating across four axes — not just capital.

"You care about the investor's name brand because their credibility helps you in terms of attracting employees and customers and so on. You care about their strategic advice and their value. You obviously care about their money, their capital, and you care about the network that they have." [00:25:41]

Private vs. Public Company — A Fundamentally Different Game

Jagdeep offers a candid inversion: public companies have more cash but less freedom to spend it, while private companies are cash-poor but operationally liberated.

"Before you're public, you're cash poor, but you don't care about the P&L because it's just paper losses... Once you go public, you raise a lot of money. Now you have a very strong balance sheet, but you can't really spend it because you have to be P&L positive... I don't know if I love running public companies, but I do love building companies. And I think a lot more of the building happens while you're still private." [00:15:25]

2. Contrarian Perspectives

If Everyone Agrees With Your Idea, It's Already Too Late

Most founders seek validation and treat universal enthusiasm as a green light. Jagdeep inverts this entirely.

"If everybody says it's a great idea, it's probably already too late because it's already priced in, if you will... You want an idea where it's different from what everybody else thinks is the right way to do it. And then you want to listen to the criticisms and see if you agree with them. Don't get swayed by the emotional aspect of the criticism." [00:10:43]

Customer Enthusiasm at Concept Stage Matters More Than Product-Market Fit Later

Conventional wisdom says validate with a product. Jagdeep argues the strongest signal comes before a single line of code is written — and that mild interest ("come back when you have a product") is disqualifying.

"Before we even spend a dime on building anything, I like to engage with the potential customers and say, look, here's what we're thinking... If they say, yeah, it's interesting, come back when you have a product, that actually is not a good answer. The answer I want to hear is, well, this is so compelling that we want to help you get this to market." [00:04:23]

Selling a Company Early Is a Failure of Vision, Not a Success

Most startup ecosystems celebrate early acquisitions as wins. Jagdeep frames the Latera sale — $550 million at 10 months — as a mistake, because the company went on to generate billions for Sienna.

"Vinod was the one guy who saw the potential of what the company could become and urged us not to sell... Post the sale, the company ended up generating, you know, literally billions in revenue for Sienna. And there was no doubt in my mind that it could have been a great independent company." [00:13:37]

Venture Capital Is a Trap for People Who Like to Go Deep

The conventional wisdom is that moving from operating to VC is a prestigious step up. Jagdeep found it unsatisfying precisely because of its breadth.

"The venture capital lifestyle, if you will, is one where you're involved in multiple companies, but not very deeply in any one... For me, I tend to want to go deep into whatever I do. And I wouldn't be happy just serving on a board and giving advice to the founders or CEO and then going away." [00:18:17]

Execution Planning Is Risk Reduction Planning

Most people think of execution plans as milestone roadmaps. Jagdeep reframes them as ordered risk registers.

"Thinking of your execution plan as being a risk reduction plan... you can literally rank order the lists and start addressing them one by one. And that becomes your execution plan." [00:32:17]

3. Companies Identified

Rhoda AI AI-driven robotics company focused on bringing general-purpose robots from the laboratory into real-world manufacturing and logistics environments. Founded by Jagdeep Singh as his seventh company. Why mentioned: It is Jagdeep's current venture and the central subject of the episode; raised a $450 million Series A with backing from Vinod Khosla, John Doerr, and Bill Gates.

"We have raised $450 million in the Series A. So it's enough capital. And with that, we're attacking the problem now." [00:24:48]

Invenera Photonics/networking semiconductor company founded by Jagdeep Singh; taken public at a $1.2 billion IPO valuation. Why mentioned: Used as a case study for the transition from private to public company operating challenges, and for the depth of commitment required in deep tech.

"Invenera was almost 10 years... they were certainly kept my interest." [00:20:04]

QuantumScape Solid-state battery company co-founded by Jagdeep Singh; also taken public. Why mentioned: Cited as a 14-year deep tech commitment and example of the long time horizons required for material science companies.

"QuantumScape was 14 years... the technologies in both companies were deep tech, kind of difficult technologies." [00:20:04]

Latera Networks Early networking company founded by Jagdeep Singh; acquired by Sienna Communications for $550 million approximately 10 months after founding. Why mentioned: Primary case study for why founders should not target M&A exits, as the company went on to generate billions in revenue for Sienna post-acquisition.

"The company ended up generating, you know, literally billions in revenue for Sienna. And there was no doubt in my mind that it could have been a great independent company." [00:13:37]

Airsoft Jagdeep's first company, focused on network protocol optimization for low-bandwidth, high-latency wireless networks; provided Jagdeep financial independence at age 29. Why mentioned: Origin story of Jagdeep's entrepreneurial career and first proof of his four-part framework.

"While I was doing my master's degree in computer science at Stanford, one of the ideas was this notion of improving the network protocols to allow them to work on very low bandwidth, high latency networks." [00:07:04]

Rax Room (likely Nuvera/Lookout or similar — context suggests a company sold to Google for $1 billion) A company on whose board Jagdeep served as executive chairman; sold to Google for approximately $1 billion. Why mentioned: Cited as a venture Jagdeep was involved with during his QuantumScape tenure, immediately before founding Rhoda.

"You were also doing stuff on the side, like Rax Room, for example. You started through there where you were the executive chairman that sold to Google for a billion." [00:21:01]

Sienna Communications Networking equipment company that acquired Latera Networks. Why mentioned: Acquirer in the cautionary tale about selling too early; went on to generate billions from the acquired technology.

Khosla Ventures Prominent Silicon Valley venture capital firm founded by Vinod Khosla; Jagdeep spent time there as a venture partner. Why mentioned: Jagdeep used his stint there to understand the VC mindset, ultimately confirming he preferred operating over investing.

"I did realize that it wasn't what I wanted to do." [00:18:17]

Sun Microsystems Early employer of Jagdeep Singh before he became a full-time entrepreneur. Why mentioned: Part of the early career context where Jagdeep was developing ideas in notebooks while still employed.

Hewlett-Packard Early employer of Jagdeep Singh. Why mentioned: Same context as Sun Microsystems — the period during which Jagdeep maintained his idea notebooks.

4. People Identified

Vinod Khosla Founder of Khosla Ventures; formerly at Kleiner Perkins. Why mentioned: Uniquely saw the long-term potential of Latera Networks when all other investors wanted to sell; drove over on a Saturday to personally urge the management team not to accept the Sienna acquisition. Also an investor in Rhoda AI.

"Vinod was the one guy who saw the potential of what the company could become. And he urged us not to sell... He maybe drove over on a Saturday to try to convince the management team not to sell." [00:13:37]

John Doerr Legendary venture capitalist at Kleiner Perkins. Why mentioned: Named as one of the high-profile investors in Rhoda AI's $450 million Series A.

"People like Vinod Khosla, but also people like John Doerr, people like Bill Gates. These are all people that have all invested in the company." [00:24:48]

Bill Gates Co-founder of Microsoft; prominent technologist and philanthropist. Why mentioned: Named as a direct investor in Rhoda AI's Series A, alongside John Doerr and Vinod Khosla.

"People like John Doerr, people like Bill Gates. These are all people that have all invested in the company." [00:24:48]

Steve Jobs Co-founder of Apple. Why mentioned: Jagdeep credits a Time magazine cover featuring Jobs as the singular inspiration that set him on the entrepreneurship path at age 15.

"I remember, you know, when I was still in high school, I think seeing an episode, an issue of Time magazine with Steve Jobs on the cover. And that was pretty inspirational. And I felt like that's the path I wanted to pursue." [00:00:56]

5. Operating Insights

Idea Validation as a Sequential Filter, Not a Single Moment

Jagdeep kept physical notebooks for years, running every idea through the same structured checklist before acting. The process was deliberate elimination, not spontaneous inspiration.

"I had these notebooks I would keep, where I wrote down the ideas... Here's the concept. Here's the problem I think is a problem. How big is this problem? Do people really care about this problem? Can I solve it? What are some different ways to solve it? Can I convince other great people to join me in this effort? Can I actually talk to some customers to see if they really care about this? And then I would rule out the idea." [00:07:04]

Recruit by Forcing People to Confront Their Own Opportunity Cost

Rather than generic "sell the vision" advice, Jagdeep frames recruiting as a specific psychological challenge: you must convince already well-compensated, high-performing people to consciously give up the second-best opportunity in their lives.

"You've got to be able to convince the best people in the world, people who have great jobs working at the best companies in the world, to leave those great jobs at great companies and come join you in your vision... You need to be able to convince these incredible people who are very well taken care of, wherever they are, to leave what they're doing and drop the opportunity cost of the second best opportunity to come join you." [00:30:22]

Execution Planning as Rank-Ordered Risk Reduction

Rather than building milestone-based roadmaps, Jagdeep builds execution plans by listing every possible failure mode, ranking them by probability and magnitude, and systematically addressing them in order.

"You can literally rank order the lists and start addressing them one by one. And that becomes your execution plan. It's thinking of your execution plan as being a risk reduction plan." [00:32:17]

What Investor "Buy-In" Actually Means — Belief Plus Paranoia

Jagdeep makes a precise distinction between investors and team members who are enthusiastic yes-men versus those who are genuinely bought in. The latter combine belief with active risk identification.

"When I say people who are bought in, I don't mean people who are just yes men... Once you buy into the vision, then it all becomes about, well, what are the risks in executing this vision? And you've got to have people who, along with you, are willing to brainstorm everything that can go wrong." [00:31:20]

6. Overlooked Insights

The Transformer Paper Author Was in the Room and No One Knew What It Was

Jagdeep casually mentions attending a Stanford class in 2018 where a Google author of the original Transformer paper presented it in person — and nobody in the room grasped its significance. This is a remarkable data point about how even expert communities systematically miss paradigm shifts as they happen, and it directly validates Jagdeep's broader thesis about contrarian insight.

"I took my first class in AI back in 2018... The Transformer paper had just come out from Google. And one of the authors actually came into the class to present the paper. And nobody knew at the time how big this was going to be, that this was going to lead to ChatGPT and the whole revolution of AI." [00:21:57]

The investment implication: the people closest to a breakthrough technology are often the last to understand its full scope. Founders and investors who maintain a wide-angle view across domains — rather than deep specialization in one — may be better positioned to recognize the next equivalent moment.

Rhoda AI Is Vertically Integrated Across AI Model, Hardware, and Sales — A Deliberate Strategic Moat

In a single sentence that received no follow-up discussion, Jagdeep reveals that Rhoda is not building software on top of existing robotics platforms. It is building its own AI model, its own robotic hardware, and selling directly to customers — a full-stack vertical integration strategy that mirrors what made Tesla and Apple defensible, and which would be extraordinarily capital-intensive and difficult to replicate quickly.

"We knew this company was going to require a really solid balance sheet. This is not for the faint of heart. We wanted to do our own AI model. We wanted to do our own AI hardware, robotic hardware, and, you know, sell the customers directly." [00:23:51]

This vertical integration, funded by $450 million at Series A with backers of the caliber of Khosla, Doerr, and Gates, suggests Rhoda is positioning for a winner-take-most outcome rather than a niche application play — and the Series A size makes far more sense when viewed through this lens.