Hire for Agency (Bouaziz's 7 Questions)

Filter, do not develop

Ask a job candidate to estimate California's state budget for healthcare, and walk through the method with them. Alex Bouaziz does not care what number they land on. It is one of seven questions the Deel founder built into his hiring process, each aimed at a single trait he holds a remote company cannot train into someone after the fact: agency. The framework rests on a premise about elite institutions: "The key to SpaceX or Harvard is not that they produce high-quality people. It's that they filter for high-quality people. The best founders build the correct systems to attract them."¹ The operating constraint is that a remote company has no office in which to supervise anyone, so the interview must select for high agency, high intensity, self-discipline, and a desire to work hard and move fast, because nothing downstream works on people who lack those traits.

The seven questions

Bouaziz lists seven questions, each aimed at one disposition:

1. "If you could clone yourself as a 20-year-old, what career path would you send him along?" for self-awareness and drive.
2. "Tell me one topic you went from zero to ready to teach in less than six months." for learning speed.
3. "What is California's state budget for healthcare? Walk me through how you'd estimate it." for first-principles thinking, where the process matters and not the answer.
4. "Explain something you understand deeply but most people find boring or dumb." for work ethic and judgment about what matters.
5. "Is there something you've changed your mind about in the past two years?" as an ego check and a test of processing new information.
6. "Have you ever wanted to do something, lacked the resources, and did it anyway?" for agency directly.
7. "What is the best thing you've built that no one asked you to build?" for intrinsic motivation and initiative.

Bouaziz pairs the questions with a meta-move: no separate culture document. Deel states the intensity, expectations, and communication norms plainly in the interview itself, on the logic that telling the truth about the job's demands attracts challenge-seekers and repels the rest before day one. He also holds that the filter requires practitioner interviewers, summarized as being unable to spot a great engineer without being one.

Why the framework hangs together

Bouaziz's argument is that most founders treat "hire for agency" as a vibe, and that these questions operationalize it: each targets one measurable trait, and questions six and seven in particular have no strong answer unless a candidate's life actually contains the evidence, which resists a rehearsed interview story. In his framing the filter is the front gate of a coherent remote system: select self-directed people first, then manage them by outcomes rather than observation, and let intensity self-select the rest. Removing the filter, he argues, leaves the rest of the system selecting against the company.

The disposition screen here is a companion to the shape-of-talent question addressed by hire for spikes and the founding-team framing of first 20 hires seal your fate, which shares the skill-times-hunger instinct extended into a repeatable interview loop. Bouaziz notes an internal tension of his own: some strong remote performers, including a hire he cites who interviewed unusually, might do poorly on exactly the estimation and explanation questions, and a filter-do-not-develop stance concedes that the company will not manufacture agency internally.