Every Friday morning, while you're checking email, he's in a Brooklyn cafe reading AI research papers that'll tell him which billion-dollar bet to make next.
Before he was predicting the future from a cafe table, Singh was slinging salads. At Georgetown, he co-founded The Hilltoss, a student-run restaurant serving smoothies and salads under The Corp, an entirely student-owned company. He handled everything - operations, purchasing, brand development, store design. The venture opened in 2014 in the Healey Family Student Center. Students wanted healthier options. Singh gave them healthier options. No mystery.
His path wasn't typical. Georgetown to Nyca Partners, a fintech-focused venture fund where he cut his teeth on companies like Fidel, Brigit, Catch, Blend, SentiLink, Ribbon, and Revolut. Then Brigit hired him to lead Business Development and Strategy. He helped launch their credit builder product and programs to support consumers through the pandemic. Operator experience, not just PowerPoint and handshakes.
In 2021, Andreessen Horowitz came calling. Singh joined as a partner on the fintech team, investing at the intersection of fintech, consumer social, marketplaces, commerce, and healthcare. He led investments in Sardine, Carry1st (Africa's leading game publisher, $20M round), and Adaptive. His track record spoke: Brigit got acquired for $500M. Revolut hit $45B valuation. The Rounds kept growing. He was good at this.
Singh's investment thesis rests on Richard Sutton's "bitter lesson" - the idea that scale and compute always beat specialization. Most AI founders are building specialist tools the same way they built SaaS products a decade ago. Singh thinks that playbook is dead. He sees two paths to durability: build the infrastructure layer (compute, data, energy, security) that keeps models scaling, or invent entirely new workflows that only AI makes possible. Everything else is noise.
Singh practices discipline like a monk. Every Friday and Saturday morning, he's at the same Brooklyn cafe. Only his iPad. Phone stays home. Laptop stays home. He deep-dives into bookmarked AI research papers and articles. No distractions. No checking Twitter. No responding to urgent emails that aren't actually urgent.
This isn't performative productivity theater. This is how he spots patterns. This is how he identified the fraud problem before everyone else saw it coming. This is how he knows that the length of tasks generalist models can achieve is doubling every seven months. This is how he built conviction that commercial infrastructure - not technical infrastructure - will create the most value in this AI wave.
That quote captures Singh's worldview. He's not betting on the smartest algorithm or the most elegant solution. He's betting on scale. He's betting on infrastructure. He's betting on the boring stuff that makes the exciting stuff possible.
In 2018, while investing in billion-dollar startups, Singh co-founded Creature/Of, a sustainable streetwear brand. Not a side hobby. A real company. The concept: modular fashion platform with customizable denim. Customers pick their fit, wash, size, and length. Made-to-order only. Zero inventory. Zero waste. Own the entire supply chain. Prices range from $175 to $195 depending on what you choose.
Why would a venture capitalist launch a fashion brand? Because Singh isn't just an investor. He's an operator. He builds things. The Hilltoss taught him that. Creature/Of taught him supply chain, sustainability, customer experience, brand positioning. When he evaluates founders, he's not just reading a deck. He's lived it.
Singh distinguishes between technical infrastructure (making products possible) and commercial infrastructure (enabling 10x user experiences). He's backing the latter. Plaid was commercial infrastructure for fintech - easy bank account access. Now he's looking for equivalents in AI.
Worldbuild's investments span AI infrastructure, fintech, and consumer. Each bet reflects the same philosophy: build what models need to scale, or build what only models make possible.
Singh speaks four languages - English, Spanish, Hindi, and Punjabi. That multilingual perspective helped him spot global fintech opportunities others missed. He holds FINRA Series 7 and Series 63 certifications. He's Managing Partner at GoMassive alongside Worldbuild. He has 8,000+ LinkedIn followers and 6,800+ on Twitter.
In 2025, Singh left a16z to found Worldbuild. Not a typical venture fund. A thesis-driven investment firm partnering with creative technologists. He calls them worldbuilders. People building the post-software era. People who understand that the homogeneous software era is over. People who aren't retrofitting AI into existing workflows but inventing workflows that couldn't exist without AI.
His Substack, launched in February 2023, has hundreds of subscribers. He writes about commercial infrastructure for AI, influence-led software, compute absorption. In December 2025, he published "Two Ways to Win in the Post-software Era" on Every. The piece crystallized his thesis: most AI founders are making a critical mistake by ignoring Sutton's bitter lesson. They're building special tools that believe they can beat base models with specific workflows. But base models are becoming increasingly capable. Generalist systems are eating specialist systems. The math is unforgiving.
San Francisco Compute Company is a Worldbuild portfolio company that started as an audio model company, bought a year-long GPU cluster, realized they couldn't get out of the lease, and ended up building a cloud almost by accident. That's the kind of story Singh loves - unexpected pivots, capital efficiency, building what the market needs rather than what the pitch deck promised.
Browserbase gives agents headless browser infrastructure to navigate complex websites reliably. Interfere is a YC S25 company where Worldbuild led the pre-seed. Flower Computer Company is building something in the AI stack. Aetherflux is working on energy. Deepshard raised a seed round with Singh's backing. Each investment reflects the same pattern recognition: what do models need to get better? What workflows only become possible because AI exists?
Singh doesn't hide the ball. He's explicit about his methodology. Read research papers. Identify structural shifts. Find companies building those shifts. Write checks. Help them win. Repeat.
When everyone chased crypto, Singh spotted fraud. When everyone chased SaaS, Singh spotted infrastructure. When everyone chases specialist AI tools, Singh backs generalist model infrastructure. This is second-order thinking. This is what separates good investors from great investors. Great investors see around corners. They see consequences. They see cascades.
His essay on commercial infrastructure for AI identifies four emerging needs: authorization (Anon), domain-specific context, headless browser infrastructure (Browserbase), and memory systems. He outlines six characteristics required for commercial infrastructure companies: exponential market growth, emergence from technology innovation or regulation, criticality to the product's core function, innovative early customers, technical differentiation, and network effects. This isn't hand-waving. This is a framework.
Singh's Instagram handle is @__sumzy__ with a globe emoji. His Twitter @sumeet724 has been active since August 2010. He's been at this for over a decade. He has recommendations on LinkedIn praising his effectiveness as an intern - "quickly acknowledged as a valued and focused asset." He worked on private equity research, financial modeling, marketing materials. He built the foundation.
He helped develop "Hoya's Guide to Georgetown," a digital student guide. He worked on "Snap Some Shots," a customer engagement initiative at The Corp. These aren't resume bullets. These are projects. Initiatives. Things he shipped. Evidence of a pattern - Singh builds things. Always has. Always will.
His multilingual capability isn't decorative. English, Spanish, Hindi, Punjabi - these languages represent markets, cultures, perspectives. When he evaluated Carry1st, Africa's leading game publisher, he wasn't looking at it through a Silicon Valley lens only. When he backed Revolut's European expansion, he understood cross-border dynamics. Language shapes how you see opportunity.
Singh's long-term vision is clear: be the first check for worldbuilders creating durable, multi-billion-dollar outcomes in the AI-native economy. Not the safe bet. Not the consensus pick. The contrarian insight. The second-order effect. The infrastructure play everyone else overlooks because they're too busy chasing the shiny application layer.
Through Worldbuild, he's backing founders who understand that specialization loses to scale. N of 1 companies through commercial infrastructure rather than the SaaS playbook. Companies that make products 10x better, not 10% better. Companies that become critical, not nice-to-have.
That's the whole thesis. Two paths. Pick one. Execute relentlessly. Everything else is distraction.
Picture this: it's Friday morning. Singh walks into his Brooklyn cafe. He's got his iPad. No phone buzzing with Slack notifications. No laptop with 47 open tabs. Just him, his iPad, and a queue of AI research papers. He reads about transformer architectures. He reads about scaling laws. He reads about emergent capabilities. He connects dots. He sees patterns. He builds conviction.
By Saturday afternoon, he knows something you don't. By Monday, he's taking a meeting with a founder building it. By next month, he's writing a check. By next year, that company is announcing their Series A. By the year after, they're the infrastructure layer everyone depends on. This is how it works. This is how Singh works.
From The Hilltoss salad bar to Worldbuild's infrastructure bets, the pattern holds. Build what people need. Spot what's coming. Execute before others see it. Scale beats specialization. Infrastructure outlasts applications. Boring beats flashy. Every time.