My Co-Founder and I Have Worked Together for 12 Years. Here's What We've Learned.

We've shipped products at Loot Crate during their peak, scaled Ring Neighbors to 30 million users at Amazon, and now we're building Vora IQ. Twelve years. Still going.

The number one reason startups fail isn't bad ideas or weak markets. It's co-founder conflict. Y Combinator says it constantly. Investors ask about it in every pitch meeting. And most founders completely underestimate how hard it is to build something with another person under pressure.

Jeff Leung and I have been through enough together that I can tell you what actually works. Not theory. Not frameworks from a business book. Just what we've learned by doing it for over a decade.

How it started

Jeff and I met at Loot Crate back when the company was in its explosive growth phase. For anyone who doesn't remember, Loot Crate was a subscription box company for geek and gaming culture that scaled fast and became one of the biggest names in the subscription commerce space.

We were both early. We both saw what hypergrowth looks like from the inside, the good and the chaotic. We shipped features, navigated scaling problems, and learned how each other operates under pressure. That last part turned out to be the most important thing we ever learned.

After Loot Crate, we both ended up at Amazon working on Ring. I led design on the Neighbors app. Jeff was on the backend and infrastructure side. We helped scale the platform from launch to 30 million monthly active users. Different roles, same mission, same pressure.

By the time I came to him with the idea for Vora IQ, we didn't need to have a "should we start a company together" conversation. We'd already been building together for years. The startup was just formalizing what already existed.

Why it works: complementary, not identical

The biggest mistake founders make when choosing a co-founder is picking someone who thinks like them. It feels comfortable. You agree on everything. Meetings are pleasant. And nothing gets challenged.

Jeff and I are wired differently. I'm the product and design mind. I think in user experiences, visual systems, and AI architecture. Jeff thinks in infrastructure, scalability, and systems that don't break at scale. I push for speed. He pushes for stability. I want to ship the feature. He wants to make sure it doesn't crash at 10,000 users.

That tension isn't comfortable. But it's productive. Every decision gets pressure-tested from two completely different angles before we commit. The product is better because we don't agree on everything.

The rules we don't talk about but always follow

We've never written these down. But looking back, these are the patterns that have kept us aligned for 12 years.

Clear lanes. We don't step on each other's work. I own product, design, frontend, and AI agent architecture. Jeff and the team own backend, infrastructure, and API systems. When decisions need to be made in someone's lane, that person has final say. When it crosses lanes, we talk it through. But we never have the "who's in charge of this" argument because the lines are drawn.

Disagree, then commit. We argue about product decisions regularly. Sometimes I'm wrong. Sometimes he is. But once we make a call, we both commit fully. No passive-aggressive "I told you so" if it doesn't work. No undermining the decision after it's made. You fight for your position, you hear the other side, you make the call, you move.

Separate the work from the relationship. After 12 years, you know each other well enough to know when someone is stressed, tired, or having a bad day. We've learned to call that out directly instead of letting it bleed into product decisions. "You're not pushing back on this feature because it's a bad idea. You're pushing back because you're fried this week." That kind of honesty only comes from trust built over time.

Protect the energy. Building a startup is a marathon. We both know each other's limits. We know when to push and when to back off. There have been stretches where one of us carried more weight while the other dealt with personal things. That balance only works if both people are willing to be honest about where they're at.

What I'd tell founders looking for a co-founder

Don't optimize for skills. Optimize for trust. Skills can be learned or hired. Trust takes years to build and seconds to destroy.

Work together before you start a company together. Build a side project. Consult on something. Volunteer for a shared cause. See how the other person operates when things get hard, when deadlines slip, when the plan falls apart. That's the real audition.

Have the uncomfortable conversations early. Equity splits. Decision-making authority. What happens if one person wants to quit. What happens if the company fails. These conversations are awkward at the beginning and catastrophic if you wait until they're urgent.

Find someone who makes you better, not someone who makes you comfortable. The best co-founder relationship isn't a friendship. It's a partnership where both people are invested in each other's growth and the mission's success. The friendship is a bonus.

Why this matters for Vora IQ

The way Jeff and I work together directly shaped how Vora IQ's agents work together. Specialized roles with clear responsibilities. Shared context across the system. Disagreements resolved by data, not ego. The product architecture mirrors the co-founder relationship that built it.

When users ask why the agents feel different from other AI tools, this is the answer. They were designed by two people who've spent a decade learning what real collaboration looks like.

See the team →

---

Sam Altman predicted the 1 person billion dollar company. We built the operating system to make it real.