How We Run a Venture Studio Across Two Continents_

How We Run a Venture Studio Across Two Continents

People ask me this question more than any other: "How do you actually make it work?"

They're not asking about the business model. They've read about venture studios, they understand the economics, they get the thesis. What they want to know is the mechanics. How does a studio operate when the team is split between Sao Paulo and Miami? How do you maintain culture across continents? How do you prevent the thing from fragmenting into two separate operations that happen to share a name?

These are fair questions. I've asked them myself, many times, usually at 11 PM when something has broken and the person who can fix it is asleep in another hemisphere.

Here's what I've learned.

Why Two Cities

The dual-continent model wasn't a philosophical choice. It was a practical one, driven by three constraints.

Talent geography. The engineering talent we need — people who can build across the full stack, who understand AI-native architectures, who are comfortable with ambiguity — exists in concentrated pockets. Sao Paulo has one of the deepest engineering talent pools in the Americas. The city produces more software developers than most countries. But the US market, US capital, and US networks require physical presence and cultural fluency that you can't fake from 5,000 miles away.

Market access. AI Gens builds ventures that serve both Latin American and US markets. Having operational presence in both means we understand both customer bases at a visceral level, not through market research decks. When we're designing HOST360's hospitality product, the insights come from walking hotel corridors in Sao Paulo and Miami, talking to property managers in Portuguese and English, understanding the regulatory nuances of each jurisdiction firsthand.

Time zone alignment. This is the underappreciated advantage. Sao Paulo and Miami are one to two hours apart, depending on daylight saving time. That's practically the same time zone. Compare this to a US-India setup (10-12 hours apart) or even US-Europe (6-8 hours). We get the cost and talent advantages of a Latin American team with virtually no time zone penalty.

Physical Space as Serendipity Engine

I'm going to say something that sounds contradictory given that we operate virtual-first: physical space matters enormously. Just not in the way most people think.

The Av. Paulista penthouse serves as AI Gens' headquarters. It's not a place where people commute to sit at desks and answer Slack messages — they can do that from anywhere, and they should. It's a serendipity engine: a physical environment intentionally designed to produce the unplanned collisions, impromptu conversations, and ambient awareness that drive creative problem-solving.

The best ideas at AI Gens have not come from scheduled brainstorming sessions. They've come from two people working on different ventures who overhear each other's conversations, realize there's a connection, and explore it over coffee. They've come from a design review for one product that accidentally reveals an insight applicable to another. They've come from the energy of being in a room with people who are all building something from nothing, simultaneously.

You cannot engineer serendipity over Zoom. I've tried. It doesn't work. The bandwidth of a video call is sufficient for executing a plan. It is insufficient for generating one.

So we use physical space with intention. The Paulista headquarters hosts ideation weeks, design sprints, cross-venture workshops, and demo days. People fly in, they spend focused time together, and they leave with insights and connections that fuel months of distributed execution.

Miami serves a different function. It's our US gateway — the place where relationships with American LPs, partners, and ecosystem players are built and maintained. Miami's tech scene has matured significantly, and its cultural bridge between the US and Latin America makes it the natural meeting point. When a US investor or partner visits, they come to Miami. When we need to be in the US market, Miami is home base.

How the Team Is Structured

The team structure is where the studio model's operational advantage becomes concrete.

We run two types of roles:

Shared functions serve all ventures simultaneously. Engineering, design systems, DevOps, finance, and legal operate as horizontal layers across the portfolio. A shared engineer might spend Monday and Tuesday on Apollo, Wednesday on HOST360, and Thursday and Friday on a new venture in exploration. They carry context, patterns, and infrastructure decisions across all of them.

This is not consulting — the shared team members are not external advisors parachuting into unfamiliar territory. They're embedded operators who understand every venture deeply because they helped build them all. The institutional knowledge they carry is the primary vehicle for the pattern transfer that makes studios powerful.

Venture leads are dedicated to a single venture. Each venture has a lead who owns the product vision, go-to-market strategy, and day-to-day execution. They're the "founder" of that venture within the studio context, with the autonomy to make decisions and the accountability for outcomes.

The relationship between shared functions and venture leads is the critical interface. Get it right, and you get the leverage of shared resources with the focus of dedicated leadership. Get it wrong, and you get neither — shared resources stretched too thin, venture leads frustrated by lack of dedicated attention.

We've gotten it wrong. I'll talk about that later.

The Daily Operating Rhythm

Here's what a typical day looks like across the two continents.

Morning (Sao Paulo, 8 AM / Miami, 9 AM). The engineering team in Sao Paulo starts with async standup — a written update on what they shipped yesterday, what they're working on today, and where they're blocked. This isn't a meeting. It's a structured document that anyone across the studio can read. By the time Miami wakes up (one hour later), they have full context on overnight progress.

Midday (overlap hours, 10 AM-4 PM Sao Paulo / 11 AM-5 PM Miami). This is the synchronous window — six hours of overlap where real-time collaboration happens. We protect this window fiercely. All meetings that require cross-team participation happen here. Design reviews, architecture decisions, venture lead check-ins — all scheduled within the overlap.

Afternoon (Sao Paulo, 4 PM onwards). The Sao Paulo team moves into focused execution. The conversations from the synchronous window have set priorities and resolved blockers. The remaining hours are heads-down building time, uninterrupted by meetings.

Late afternoon (Miami, 4 PM onwards). The Miami team handles US-facing activities — LP conversations, partner meetings, ecosystem engagement — during US business hours. By evening, they produce async updates that the Sao Paulo team will read the next morning.

The rhythm is simple. Async for status. Sync for decisions. Deep work in the margins. It works because the time zone overlap is generous enough to allow meaningful real-time collaboration without forcing anyone into unreasonable hours.

What 80% Virtual Gets Wrong

Here's a statistic I think about a lot: 80% of startups now operate primarily virtual. And most of the innovation literature suggests that the most creative work still happens in person.

Both of these things are true, and they're not contradictory.

The mistake most remote-first organizations make is treating physical presence as binary: either you're in the office full-time or you're fully remote. This creates a false choice between the efficiency of remote work and the creative advantages of co-location.

The studio model offers a third option: virtual-first for execution, physical for ideation. You don't need to be in the same room to write code, review pull requests, or update a financial model. You very much need to be in the same room to sketch a new venture concept on a whiteboard, debate product strategy with the nuance that only face-to-face conversation allows, and build the trust that makes distributed execution possible.

We've found that 4-6 days per month of intentional co-location produces about 80% of the creative benefit of full-time co-location at about 20% of the cost and disruption. The key word is intentional. These aren't "come to the office and sit at your desk" days. They're structured around specific activities that benefit from physical presence: ideation workshops, design critiques, strategy sessions, demo days.

The generic coworking model — open floor plans, communal coffee, hope for organic collisions — doesn't produce these benefits. Specialized innovation hubs with intentional programming do. Our Paulista space is designed for this: flexible rooms that reconfigure for different activities, whiteboards on every wall, areas for focused work and areas for loud debate.

What Works

Let me be specific about what actually works in our model.

Cross-venture pollination. Having the same engineering team work across multiple ventures produces insights that no standalone startup could generate. The Apollo team's experience with agentic architectures directly accelerated HOST360's development. Design patterns validated in one venture become starting points for the next. This is the intellectual capital thesis in action, and it's real.

Talent density over talent quantity. Because we share senior functions across ventures, each venture gets access to more experienced people than it could afford independently. A pre-seed startup can't hire a senior architect, a seasoned designer, and an experienced finance lead. A studio venture gets all three because they're shared across the portfolio.

Cultural coherence. Having a physical home base — even one that people visit rather than commute to — creates a cultural anchor. The Paulista headquarters is where AI Gens' identity lives. People who visit absorb the culture, form relationships, and carry that back to their distributed work. It's a cultural container more than an office.

The Sao Paulo-Miami corridor. The near-identical time zones, the cultural bridge between Brazil and the US, the complementary talent pools and market access — this specific geographic combination works better than any other dual-continent setup I've seen.

What Doesn't Work

I promised honesty. Here it is.

Shared resources during venture sprints. When a venture is in a critical sprint — a launch, a major feature push, a fundraising crunch — shared resources create tension. The venture lead wants 100% of the engineer's attention. The studio needs that engineer on another venture too. We've had real friction here, and the resolution is never clean. We've learned to anticipate sprints and pre-allocate resources, but it's still the hardest operational challenge we face.

Async-only decision making. Some decisions can be made asynchronously. Most important ones can't. The nuance, context, and back-and-forth required for consequential decisions — pivot or persevere, build or buy, hire or wait — needs real-time conversation. We've made bad decisions when we tried to resolve complex questions over Slack threads. Now, any decision with significant consequences gets a synchronous meeting, no exceptions.

Onboarding remotely. Bringing a new team member into a studio is harder than bringing them into a single startup. They need to understand multiple ventures, the relationships between them, the shared infrastructure, and the studio's operating principles. Doing this entirely remotely produces superficial understanding. We now require all new team members to spend their first two weeks at the Paulista headquarters, working alongside every venture team.

Maintaining energy in distributed teams. The energy of a physical studio — the buzz, the visible progress, the ambient motivation of being surrounded by builders — is hard to replicate remotely. We've tried virtual social events, remote team rituals, and various digital culture initiatives. Some help. None replicate the real thing. The best mitigation is regular in-person gatherings, which is why we bring the full team together every six to eight weeks.

What We'd Do Differently

If I were starting AI Gens from scratch today, here's what I'd change.

More structure around resource allocation from day one. We learned the tension between shared resources and venture needs through painful experience. A formal resource allocation framework — with clear priorities, sprint-level planning, and explicit trade-off conversations — should be in place before the first venture launches, not after the first conflict.

Earlier investment in async tooling. We underestimated how much the quality of async communication tools affects distributed execution. The difference between a well-structured knowledge base and a chaotic Slack history is the difference between a team that can operate independently and one that constantly needs synchronous check-ins. We should have invested in proper documentation, decision logs, and async update systems from month one.

Dedicated space in both cities earlier. We operated out of temporary spaces in Miami for too long. Having a dedicated physical presence in both cities — even a small one — signals commitment and provides the serendipity engine that drives creative work. The ROI on physical space is high when used intentionally; we just should have committed to it sooner.

The Model, Simplified

Here's the operating model in a few sentences:

Build the team in Sao Paulo, where the talent is deep and the culture is strong. Maintain the US presence in Miami, where the capital and market access live. Use physical space as a serendipity engine — intensive, intentional, periodic — not as a daily commute destination. Run shared functions across ventures for leverage. Dedicate venture leads for focus. Synchronize during the generous overlap window. Execute async the rest of the time.

It's not perfect. Nothing distributed is. But it's honest, it works, and it gets better as the team develops the muscle memory of operating this way.

The question I started with — "How do you actually make it work?" — doesn't have a single answer. It has a hundred small answers, refined daily, grounded in the belief that the right people, the right structure, and the right rhythm can overcome any number of miles between them.

That's how we run a venture studio across two continents. Not by eliminating the distance, but by designing around it.