Key Takeaways

  • OpenDoor, a company transforming real estate, operates with a surprisingly lean engineering team of fewer than 70 people. This low headcount defies typical tech company scales.
  • Their engineers don't just build features; they create systems designed to give non-engineers outsized leverage, amplifying overall company output.
  • CEO Kaz Nejatian insists on company-wide AI proficiency, stating that anyone at a tech company should be able to write a SQL query using tools like Claude or ChatGPT.
  • This intense focus on AI and lean engineering underpins OpenDoor's strategic shift to a market maker identity, aiming for extreme velocity and a "checkout" experience for real estate.

The Method: Max Leverage, Minimum Engineers

Imagine running a tech company of significant scale with fewer engineers than most Series B startups. Kaz Nejatian, CEO of OpenDoor, is doing exactly that. He revealed they operate with “fewer than 70 engineers,” a number he expects would "surprise most people." This isn't just a cost-cutting measure; it's a deliberate, tactical choice that reshapes the entire organization.

The core of this method lies in a fundamental redefinition of an engineer's role. At OpenDoor, engineers aren't just coding new features or fixing bugs for internal users. Nejatian explains, “Our engineers spend their time creating systems that allow other people who are not engineers to create leverage for themselves.” Think of it: a small team of highly skilled engineers builds the core machinery, then equips everyone else—from internal communications to operations—to run those machines with maximum efficiency. This pushes power and capability out to the edges of the organization.

The second, equally important pillar is a radical mandate for AI proficiency. Nejatian makes it clear: using AI isn't optional, it's a job requirement. “If you cannot, using Claude or ChatGPT or Codex or Grok, write a SQL query, you should not be working at a tech company.” This isn't just about technical roles. He points to their “head of internal coms” who “spends most of her time on Claude.” This expectation means everyone, regardless of their title, is expected to become a power user of AI, offloading routine analytical and creative tasks from engineers and enabling faster decision-making and content generation across the board. This isn't about automating people out of a job; it's about automating tasks so people can do higher-value work.

OpenDoor's strategic shift from excessive operating expenses to a disciplined, cash-flow-funded approach is directly supported by this lean, AI-driven structure. They're not trying to be a traditional asset manager; they're building a market-making machine focused on velocity and bundling services. A streamlined, AI-proficient workforce enables them to chase this ambitious vision for a simplified real estate "checkout" experience with remarkable efficiency.

Where This Breaks Down

This method isn't for every company. It relies on a few critical assumptions that might not hold true for your startup. First, it requires an extremely high bar for engineering talent. If you have fewer than 70 engineers, each one must be disproportionately effective—capable of building robust, scalable systems that are intuitive enough for non-engineers to wield. This is not easy talent to find or retain. Second, it demands a culture of relentless self-sufficiency and technological adoption from every employee. Not every team member will embrace the idea of writing SQL queries with AI, and forcing it without adequate training or buy-in could lead to resentment or errors.

Furthermore, this model might struggle in highly fragmented or non-standardized industries where data quality is poor, or AI tools cannot reliably perform complex analytical tasks. OpenDoor operates in real estate, which, despite its fragmentation, has structured data points that AI can parse. If your problem space is truly novel or heavily reliant on nuanced human judgment, a purely AI-driven leverage model for non-engineers could backfire. It also implicitly means that your engineering team spends less time on direct product features or R&D for entirely new capabilities, and more on building internal tools. This trade-off needs careful consideration.

What to Do With This

Tomorrow, identify three repetitive tasks currently handled by your non-engineering teams that could be partially or fully automated using readily available AI tools like ChatGPT or Claude. Then, mandate that the relevant team members spend two hours this week learning to use these tools for those specific tasks. Instead of asking engineers to build a custom solution, challenge your team to prototype an AI-driven workaround. Next, review your next five non-technical job descriptions. Add "Proficiency in AI tools (e.g., [Name specific tool relevant to role]) required for data analysis/content generation/customer support" to the requirements. This shifts the hiring bar and signals a new company expectation.