Building & Owning at the Same Time

I spent the first decade of my career thinking about what to build. I am spending the next one thinking about what to own.

The shift did not happen all at once.

I studied computer science. I became an engineer. I worked on large-scale machine learning at Google across Search, Cloud, and user modeling. I left to lead engineering at Sprinter Health, building technology for healthcare delivery in regulated workflows. Then I co-founded goodfin, an AI-native fintech company building investment infrastructure for accredited investors.

For most of those years, the question I optimized for was the builder's question. How do we ship faster. How do we solve harder problems. How do we make this system work in production.

Those are the right questions for an operator. They turn out to be incomplete questions for a career.

What you learn inside the workflow

The most important thing I learned in production AI was not technical. It was structural.

When you build on foundation models, you are renting your moat. The labs that supply your intelligence are also building products that compete with you. Your margin is their pricing decision. Your roadmap is their next release. The economic logic only works in one direction.

I wrote about this in detail when Cursor sold to SpaceX. The compute ceiling is real, and it is structural. The companies that survive it are the ones that own something the labs cannot replicate.

• Workflow.
• Data.
• Trust.
• Distribution.

The pattern is the same whether you are building a wealth management platform, a clinical operations product, an insurance underwriting system, or a legal AI. The durable companies do not win by having access to the best model. They win by owning a piece of the workflow the labs will not enter.

That insight reframed how I thought about my own career.

The question shifts

If the structural lesson of this era is that ownership of the workflow beats access to the model, then the same logic applies one level up. Ownership of what gets built matters more than access to the building.

This is the shift I underestimated for years. I was good at the building part. I treated ownership as a downstream consideration. Something to optimize later, after the product worked. Many builders make the same mistake.

The reason it matters now, more than in any previous software era, is that AI has changed the economics of creation.

Small teams can ship what used to require organizations of fifty. The capital intensity of building software has collapsed. What has not collapsed is the importance of who owns the result.

If five people can build what fifty people used to build, the question is no longer whether you can build it. The question is whether you can build it and own a meaningful piece of what it becomes.

Builders who do not think about ownership early end up building someone else's company. Builders who do, end up building their own.

Investing changes what you see

When I started investing, the first thing I noticed was how much faster patterns became visible.

You meet founders solving entirely different problems in entirely different industries and you start to see the structural similarities. You watch markets evolve in real time. You see which technical decisions compound and which do not. You develop a feel for where conviction is justified and where it is just enthusiasm.

The other thing you notice is how much ownership compounds.

• Relationships compound.
• Reputation compounds.
• Communities compound.
• Investments compound.

The most valuable outcomes I have seen rarely came from a single great decision. They came from a series of decisions made over years, where each one made the next one easier.

That observation became one of the foundations of how I want to spend the next decade.

434 Bayview

The name 434 comes from my first home.

434 Bayview was the address where many of the most important chapters of my life began. Looking back, what stands out is not the house. It is what emerged from it. Conversations that became friendships. Friendships that became collaborations. Collaborations that became companies.

The things that matter most usually start small. A conversation before a company. An idea before an investment. A community before an institution.

House of 434 was built around that belief. The highest-signal ideas and the strongest relationships live in private rooms long before they become visible. The best investors and operators I know spend most of their time in those rooms. The rest of the industry catches up later.

Why 434 VC

The thesis behind 434 VC follows directly from what I learned inside production AI.

The next decade of category-defining companies will be built by operators who lived inside the workflows they are now rebuilding. They will own the data, the distribution, and the trust in markets where horizontal intelligence is necessary but not sufficient. They will be smaller teams than any previous generation of software founders, with more leverage than any previous generation thought possible.

What they will need from investors is not capital alone. Capital is abundant. What they will need is partners who understand the workflow, the buyer, and the structural constraints of building AI in regulated industries. Partners who can pattern-match from inside the constraints rather than outside them.

That is the role I want to play. The fund is the vehicle. The thesis is what I learned by building.

Where this leaves me

I spent the first decade of my career as a builder. I am spending the next one as a builder, an investor, and an owner.

These roles used to feel separate. They are becoming more connected with every cycle of this technology.

The capital intensity has dropped. The team intensity has dropped. What has not dropped is the discipline of deciding what to own. That is the question I think most about now, and the question I will spend the next decade helping founders answer.