A working document on how I think hiring and management should actually be done. This is not a polished corporate handbook. It's a philosophy of how to build a company that runs on intellectual honesty, ruthless execution, and shared obsession with the mission. Great companies are philosophies. Not because of their identity, but because of the structure they carry — and that structure itself becomes the identity. You lead with that, or you don't lead at all.
Table of Contents -
Hiring
Compensation and Work Ethic
One-on-Ones and Culture
Management Structure: The RNA-DNA-Protein Model
Criticism Management
Idea Management
Evaluating and Firing
Closing Note
Hiring is the single most important operational decision a company makes. Everything downstream — culture, speed, intellectual horsepower, willingness to push hard — is set by who walks through the door. So the process needs to filter for substance, not just credentials.
Before anyone gets to a real interview, we want technical and skill diligence done up front. The application form should ask for relevant materials — projects they've actually built, a short message or story about who they are and how they got here. Keep it under 200 words. The constraint is intentional. People who can't compress their story into 200 words usually can't compress their thinking either.
Led by either the team lead or the project lead, not by HR, not by a generalist. The point of this interview is to test whether the person can actually build. Give them a real project. A feature to build. An existing feature to improve. Something with enough surface area that we can see how they think when they hit a wall. Not a leetcode puzzle. Real work, real constraints.
The final interview is where we figure out whether this person is someone we actually want to spend years building alongside. Five things matter here.
Life story. Tell us how you got to this moment. The crazy things you've done. The things you've built that nobody asked you to build. What you've quit, what you've stuck with, and why. We're looking for the texture of a real life, not a resume read aloud.
Top priority. What is the one thing in your life that you consider a top priority? If they can't answer this cleanly, they don't know themselves well enough yet. Self-knowledge is a prerequisite for working on something hard, because hard work surfaces every weakness you didn't know you had.
Short essay on the problem we're solving. Write us an essay on what we're building. Is there anything we can improve? If yes, what, and how? Keep it focused on the improvements and the process behind them — not flattery, not generic praise, not a pitch about themselves.
Talk to our users. This is the test that filters more than anything else. Talk to our existing users — not as a representative of the company, but as a curious person trying to find a solution to a problem. After the discussion, ask them how they'd rate the product on seamlessness, functionality, and as an actual solution to their problem. The candidate has to come back to us with what they learned. If they can't extract real signal from a real user conversation, they can't help us build.
Chess game with the CEO. This sounds strange but it's deliberate. A chess game reveals how a person thinks under pressure. How fast they make decisions. How good those decisions are. Whether they think about consequences and second-order effects. Whether they play attack or defense by default, or whether they actually adapt to the situation. None of this is about chess skill. It's about cognition under structured uncertainty, which is what work actually is.
Always hire people smarter than you. Then give them the plan and, for anything technical, ask them about it and genuinely value their opinions. If you're hiring people you can dominate intellectually, you're hiring backwards. The point of bringing someone in is to raise the ceiling of the company, not protect your ego.
If something can be built in a week, build it in a week. Not two. Speed is a feature of the company itself, not just the product. Slow companies die — sometimes loudly, more often quietly.
Whether you're part of it or not. There's no room for modesty or kindness as a substitute for performance. Be ruthless. Kindness in hiring and firing is often cowardice wearing better clothes. Real respect for people is treating them like adults who can handle the truth about whether they fit.
If you want obsessed people, pay them like you believe in them. Underpaying obsessed people is how you lose them to someone who's willing to recognize what they're worth. Compensation is not generosity. It's alignment.
If I'm taking sleepless nights and working my ass off to make this mission real, you should be too. This is not negotiable. Vision and execution alignment is the foundation. A "ruthless work routine" does not mean neglect your health or your family. It means: when you work, you work with full focus and full conviction. You spend time with your family, you talk, you eat, you live — nobody's stopping you. But when you sit down to change the world, the obsession has to be real. Half-obsession produces half-companies. If you don't love your work — if you don't love every word of code, every line of design, every paragraph of writing — if you're not obsessed with making it perfect and functional — don't apply here. We only want people who love what they do. Everything else is a waste of their life and ours.
One-on-ones are good. They can unravel issues, shift behavior, build trust. But they're not strategic in terms of execution speed, and they can quietly undermine the culture you're trying to build.
Here's the problem. If you want a world-class company, you need a culture that celebrates showcasing problems, learnings, approaches, and roadblocks — out loud, in front of everyone. Every problem should be treated as a story that reveals patterns about the company itself, not just the individual situation.
You cannot build that culture in private rooms. One-on-ones move those conversations behind closed doors, and the moment that happens, transparency becomes a performance instead of a habit. People learn that the real conversation happens elsewhere, and the public culture becomes a sanitized version of reality.
The opposite — open, public discussion of problems and learnings — requires real sacrifice. It requires the will to swallow your humility, sit with discomfort, and channel that discomfort into learning. Most companies refuse to do this because it's hard. That refusal is exactly why most companies are mediocre.
Use one-on-ones sparingly and for the right things. Build the culture in public.
Keep the structure simple. The principle underneath it is biological: let ideas flow through the company the way energy flows through the body. Constriction anywhere weakens the whole system.
Every team has one team leader. Teams work interconnectively — ideas flow freely between them, not just within them.
Each cluster of nearby teams sits under one managing director (MD). The MD must be intellectually fluent in every domain their teams cover. If they aren't, they get replaced. An MD who can manage process but cannot evaluate the substance of the work is a gatekeeper who blocks signal instead of carrying it.
The MD takes care of their product and segment's teams, but they also work interconnectively with other MDs and accept ideas from those teams too. This isn't optional. The whole structure depends on lateral flow, not just vertical reporting.
Multiple clusters like this exist in parallel. Smaller teams, one team leader each, one MD over the cluster.
All MDs report into one leadership — the CEO. Everyone, at every level, is working toward the same thing: making the company and the vision succeed as a whole.
No real alignment with the vision means immediate dismissal. No matter the skills. If someone doesn't feel the pull of the mission, they will eventually drag on it. The CEO's job here is to make the vision easy to understand and easy to work on — that's the precondition for holding anyone else accountable to it.
Criticism is not avoided. It's celebrated. Including criticism of leadership. If a leader is wrong, the team is expected to say so. Out loud. In public. The day this stops happening is the day the company starts dying without knowing it.
Ideas from the smallest role are considered seriously. The person closest to the problem usually has the highest-resolution data about it. Ignoring that data because of where it comes from in the org chart is just self-imposed information loss.
The criticism submission process exists for one reason: to make sure criticism is useful, not just emotionally vented. The form forces people to own a real claim instead of hiding behind vagueness.
Every criticism submission must answer:
Why this criticism? What's the actual reason behind it.
Why is it necessary? Necessity forces the critic to think about whether this matters or whether they're just irritated.
What changes if we apply your thesis? A criticism without a proposed direction is a complaint. We need to know what reality looks like if you're right.
Is your criticism viable, or is it vague? Self-assessment. Can this actually be acted on?
What did the person do that made you write this? Concrete behavior, not interpretation.
Did you try to talk to him? This is the most important question on the form. Our culture is flexible — but if you didn't try to talk to the person directly first, we can't take the criticism. Going around someone instead of to them is institutional cowardice, and we don't reward it.
Share your problems and your criticisms publicly. Do not let criticism define you. If something is wrong, work on it. Upgrade yourself. The point of criticism is not to wound — it's to surface reality faster than ego can hide it.
Ideas are cheap. Worked-on ideas are rare. The idea form exists to test whether the person submitting actually has the intellectual honesty and follow-through to be worth listening to.
Why did you choose this specific idea? What's the real reason — not the pitch version.
How many ideas have you previously submitted and actually worked on? Did it succeed? If not, what stopped it? What did you learn? This question is the most revealing. Someone who has submitted ten ideas and can't tell you what they learned from any of them is a pattern of wishful thinking. Someone who failed multiple times and can give honest post-mortems is worth betting on.
What value does this idea add in the world? Real value. Not "it would be cool if."
What's your role in it? Are you the right person to execute this, or are you just good at generating concepts? The honest answer separates vision from vanity.
How many people are already signed up for this? If no one, why?
Did you validate something? Talked to potential users, did research, built an MVP, anything. If not, what stopped you?
The point of all of this is to make your thoughts clearer about your idea and about yourself, and to create a productive loop. The form is the work, not the obstacle.
Look for operational fluency in your MDs. Not just management skill — actual fluency in product, ideas, execution, and the day-to-day work of getting things built. The bar is top-tier. Anything less and the role degrades into administration.
Build Systems for Firing — Don't Fire Into the Void
Firing should not be impulsive, and it should not be avoided either. Build a system that diagnoses before it acts.
When something isn't moving:
Ask yourself first. Is this person actually good at what they do? Be honest. Have you given them what they need? Are the conditions around them broken in a way that would defeat anyone?
Ask them. What's the reason this isn't moving? Listen for whether the answer is a diagnosis or an excuse. A real operator gives you a clear-eyed read on what's wrong, even when it's uncomfortable. Someone deflecting is a flag.
If there are flags, fire. No looking back.
If there are no flags, give an extended timeline. Make it explicit. Define what needs to move and by when. Then watch.
If nothing moves after the extended timeline, fire. No looking back.
The principle here is that mercy without diagnosis is cruelty in disguise — it keeps people in roles where they're failing without giving them a real chance to correct, and it lets the company decay around them. Real respect is honest evaluation followed by clear action.
Great companies are philosophies.
Not because someone declared a vision statement. Not because of the brand. Because the structure itself — how people are hired, how they're fired, how ideas move, how criticism is handled, how decisions get made — embodies a coherent worldview. The structure is the philosophy. And the philosophy is what people actually experience when they work there.
Lead with that, or don't lead.