One of the most interesting shifts in this wave of AI-native startups is the inversion of the typical founder profile. Historically, particularly in vertical SaaS, many companies were founded by domain experts who hired technical teams to build software for problems they had experienced firsthand. Domain knowledge was the starting point; technical capacity was layered in later.
What we’re seeing now is the reverse. AI-native companies are largely led by technical founders, specifically people who deeply understand how to work with large language models and instinctively grasp what they can do. They’re not insiders to the industries they’re building in; they’re experts in the toolset. And instead of starting with context, they hire it as early as possible. Nearly every AI-native vertical product we’ve seen is being built this way.
That inversion matters because it affects how these companies build differentiation and where they find defensibility. Founders who understand how to build with AI have a real edge. These models are powerful, general-purpose tools, and knowing how to prompt, fine-tune, and orchestrate them unlocks product capabilities that are 10x to 1000x more efficient than human labor in certain workflows. That creates magic in the user experience. It lets teams move quickly, build new interfaces, and invent entirely new categories of product. That’s differentiation.
But differentiation is not the same as defensibility. Just because a product feels magical doesn’t mean it’s hard to copy or hard to rip out. Moats haven’t fundamentally changed. Defensibility still comes from owning the workflow end-to-end, embedding deeply into customer systems, becoming the system of record, or earning trust over time. These are context-specific advantages. They don’t come from knowing what AI can do – they come from knowing why that matters in a given domain. Which edge cases break trust. What thresholds of accuracy are acceptable. How data flows through legacy systems. That kind of judgment can’t be automated like code. It varies dramatically by domain and is earned only through experience, and it’s what turns a great product into an irreplaceable one.
That’s why context still matters and why integrating it early into the company and product loop is critical. AI dramatically expands what’s technically possible, but it doesn’t tell you what’s actually useful. You still need deep domain knowledge to figure out how to translate model capabilities into product value. And as the pace of change accelerates, that translation layer becomes more time-sensitive. Founders now need to track not only model releases, but understand the downstream implications like recognizing when a new use case or workflow suddenly becomes automatable. Being early to that shift unlocks growth – new product surfaces, new wedges, and new value to your customers. While being late, even by a few months, can be costly.
We’re seeing this play out across our portfolio. Companies like Tennr and Eve didn’t start with deep industry roots. Their founders were technical, AI-native, and fast-moving. But they quickly brought in people who had lived the work to help define the product. These context hires use the product and ensure what gets built actually works in the messy, high-context workflows of a specific field.
AI is an extraordinary engine for product differentiation. But it’s not a moat. We believe the companies that endure will be the ones that embed context deeply, pairing technical fluency with domain expertise to build products customers can’t live without.
