Hiring remains the biggest topic of discussion among the founders I work with.
AI has amplified talent wars and widened the dispersion of talent, resulting in venture-backed AI startups ferociously competing for top 1% operators.
I’ve been thinking about how obsessive founders need to be with hiring ever since reading Brie Wolfson’s profile on Cursor last year:
Cursor’s secret to recruiting is to treat the atomic unit of the hiring process as a person, not a job spec.
There’s a constant stream of names flowing through the #hiring-ideas channel. Sourcing doesn’t consist of searching for relevant job titles or companies on LinkedIn and adding names to a spreadsheet for a recruiter to reach out to; it looks more like genuine curiosity about who the best people are.
Still, despite all the outbound, the acceptance rate at Cursor makes elite colleges look like summer camp. Every leader says that talent is king, but few companies actually commit to it. Cursor does. Choosing not to hire someone when the to-do list of important problems is as long as it is at Cursor is not easy. But as a smart founder friend put it, “they’re pulling the pain forward.” Cursor’s leadership team signs off on every hire, and I suspect that will be the case for a long time.
Just as everyone sources candidates, everyone closes. Occasionally, someone needs a little more coaxing post-offer. (Reminder that “looking for a job” is a notable non-requirement for being offered a job at Cursor).
In this respect, the team is relentless. Ryo Lu, former early designer at Stripe and Notion, and an Apple fanboy, was gifted an early edition Macintosh computer. Lukas Möller impressed the founders with a cold email about his love for coding and appreciation for what the team was building. Despite the founders making a recruiting trip from California to Germany, Lukas declined the offer. But as Oskar told me with a smirk, “‘No’ is often the start of the conversation.” A year later, the founders were on a plane to Germany again, and this time Lukas came back with them to SF.
I’ve always enjoyed Lenny Rachitsky’s pieces on how companies win their first 10 customers, when companies hired their first PM, and more, as they’re brimming with tactical advice that founders could action.
So, I got my Grok Bots working on scouring all podcasts, articles, and X posts made by founders or employees of companies widely regarded as having exceptional talent density, with a focus on companies founded in the post-ChatGPT era (with some exceptions).
To structure this, I’m using the structure Elad Gil applied in Chapter 3 of High Growth Handbook, focusing on recruiting, with onboarding for a future post.
Part 1. Recruiting best practices
A1. Write a job description for every role
Tactic: Scope and stack-rank before talking to anyone (Cursor)
“The alternative is treating every hire like an executive search: rigorous scoping (what does “great” mean for this specific role?), deliberate candidate mapping (who are the 50 best people in the world for this role?), and relentless pursuit (go after these 50 people and don’t stop).
Really good scoping is thinking about what are the skills and experiences you’re looking for. You stack rank these things, you get really clear objectively before you talk to any candidate, but it sets up your entire process. It sets up your sourcing strategy. It sets up your pitch
Tactic: Shorter specs built around one spike (Ramp)
We often like to write very short job descriptions. If there’s one thing that matters, go just for that. Have the interviews center around just one thing and look for extraordinary spikes.
Tactic: Values and “not wanted” stated in the JD itself
tpuf is hardcore & whimsical. we value: overstep > understep; show > tell; complexity is earned
“Long hours, high pace. Candidates must thrive under high urgency under AGI timelines approaching. Those seeking comfortable work need not apply.”
Tactic: The trait spec: slope, trajectory, depth
We hire for potential and growth trajectory, slope over intercept. We make bets on people. The goal is not zero defects, but 10x potential.
I think that the P 99 engineer has some history of having bent, like their trajectory or something to their will.
Right? Some moment where it was just, they just, you know, made the computer do what it needed to do.
We look for people that are really high agency because we just recognize that by default, if we don’t innovate and do crazy things, we’re going to die.
There’s an extreme bias for action that I try to bring in and try to encourage everybody else in the company to adopt.
The most important thing is that people care a lot… that care or preferably obsession gets you a very long way and then you do often want to have absolute superpower in some dimension… generalist brain… but be super, super good in one dimension.
Tactic: Profile: former founders
Of our initial kind of 35, I think 21 of us have founded a company before
We have an incredibly high hiring bar where we hire tons of former founders, people that have incredible experiences.
if I look at the folks on our team who have been so excellent, a lot of them are former founders. They’re folks who started businesses, persevered, ate glass, had a lot of product intuition, figured out how to channel it, and then the business maybe petered out… we’ve acquired six different fledgling financial AI startups. And so those former founders who can be galaxy-brained about the future of the product and navigate a course on what the UX should be, but also have engineering chops so they can constantly make the right decisions, is super important.
A2. Ask every candidate the same questions
Tactic: A standing question asked of everyone
#1 thing I look for in interviews is whether someone has a history of understanding 2-4 levels below the abstractions they’re working with, e.g.: 1. Can use an ORM 2. Can write good SQL 3. Understands how indexes work 4. Can tell you how the data is roughly laid out on disk
Tactic: interview the candidate’s own claimed spike
“If you’re telling me you’re great at something, I’m going to see how great you actually are. And you better be a lot more knowledgeable about it than me after doing a couple hours of research.”
“you’re saying you’re great at poker, for example. Are you a great poker player? How good you actually are? Are you a good judge of yourself? Are you aware of the spectrum of talent in that field?”
A3. Assign focus areas to the interviewers prior to the interview
Tactic: Cross-functional bar raisers and polymath screens (Decagon)
In terms of how we maintain the hiring bar, we actually have a group of bar raisers internally that do the final interview. So once you’ve gotten the thumbs up from everybody, one of the bar raisers interviews you and we’ve specifically picked out the bar raisers as people within Decagon and we’re like, “Hey, these are the people that really kind of hold our values. They’ve demonstrated it over time and they know what they’re looking for amongst new folks.
Tactic: Values or culture as an assigned focus area
We have a values interview for all candidates when they come on site.
A4. Work Product Interviews
Tactic: Multi-day work trials (Cursor, Lovable)
research has shown that work samples are the highest predictor of success at a company, but yet most companies are still spending a lot of time doing one-to-one across the table assessments. And so I think we are famously known for doing long onsite experiences for some of our roles where they do a project and they get work to side by side. So that gives us a very strong signal. It also gives them the candidate tremendous signal.… And I’m a big fan of the power of self-selection.
the go-to-market team does a challenge where they work through a challenge together, a challenging customer, things like that… Product does a similar project type together as well.
I pretty much always have people join the work simulation for at least a day, often a full week.
Tactic: Short-form AI-native builds (Cognition, Sierra)
our whole interview process… is basically just having people build their own Devin in eight hours, and seeing how far they get with it… their own full end-to-end agent in eight hours, or six hours
Things like memorizing facts about a particular language… or being optimally efficient on a leetcode-style algorithm question, are far less important today.
We removed our coding and algorithms interviews and replaced them with an AI-native onsite.
The interviewer steps out and the candidate brings the idea to life over 2 hours, using the AI tooling and frameworks of their choice.
Tactic: Impossible-task tryouts (Etched)
We called him up a couple of years ago and said, hey, can you come check out what we’re doing? We need some help on the platform side. He comes for a week, and we say, can you build a cold plate this week? If you asked any thermal engineer anything like that, they would think you’re totally naive — these things take months to do… And he built a contraption in a week that actually de-risked a pretty key power question we had.
Tactic: Proof of past work instead of a CV (Ramp)
I’m less interested in what is the resume. I’m far more interested in proof of work. It can be an award. It can be something you built.
A5. Candidate scoring
Tactic: Default reject: no means no, and someone must fight for a yes
I have a document called traits of the p99 engineer and it’s a bullet point list and I look at that list after every single interview… I end with um some version of i’m going to reject this candidate completely irregardless of what the discourse was because I want to see people fight for this person because the default should not be we’re going to hire this person the default should be we’re definitely not hiring this person
even now including for ICs, we review hiring packets… Have we found a clear champion within Decagon? We… hire on the one, two, three, four scale. And if you just have threes all across the board… there’s no clear champion… we shouldn’t be making them an offer.
Tactic: score strengths, not the absence of weakness (Sierra)
“We’re hiring for strengths, not just an absence of weakness… Our debriefs have shifted from ‘should we hire this person?’ to ‘where would this person thrive, and how do we support them?’”
A6. Move fast
Tactic: Speed once the candidate is ready (Cursor)
Great. You’re going to talk to them in an hour… when we go, we go”
A7. Check candidates’ references
Tactic: References as a sourcing channel (Turbopuffer)
one of our recruiters convinced a reference to interview during a call about a different candidate. they just accepted. this is what p99 recruitment looks like
Part 2. Scaling a recruiting organization
B1. Early days: Your team as recruiters
The Cursor team hired their first 80 people, roughly 80 people on their own. They know how to do it. They know how hard it is, which is great, but also they love it and they’re in it and they want to do it.
“for our first 10 people, we were just so patient and disciplined about finding some of the best people in the world… And I think that that initial talent density shaped so much of what the rest of the org looks like as you scale it up.”
Tactic: The whole company sources: name channels and network mining
We have a Slack channel called Talent Spot for people to basically say, ‘Hey, I connected with this person… on X… I think they’re great talent…’
“It’s almost like social credibility in the company to be able to surface someone here is great… This is non-recruiting folks in the company dropping in ideas and then the company rallying behind it.”
Tactic: Ask your network targeted questions, not “who’s the best?”
The mistake that most companies make is they’ll say, “Hey, Lenny, who’s the best product engineer you know?” That is the worst question you can ask. The better question to ask is, “Hey, Lenny, of all the product engineers you’ve worked with, who stands out to you as the person who’s most collaborative with designers or who can translate a framework into a product better than anyone you’ve ever seen?
Tactic: Long-cycle pursuit: “no” is the start of the conversation
You really got to keep at them. When you hear no from somebody who really is the best in the world, that really means, hey, you should come back when you have a few more milestones proven out.
B2. Initial scaling: The in-house recruiter
Tactic: Hiring managers own the hire; recruiters are a “confidence engine”
“you can’t say, ‘Hey, it’s the most important thing and I’m going to outsource it,’ or, ‘I’m going to hand it over to a recruiting organization to do.’”
Tactic: Closing: caring, meals, home games and personal gestures
I think the second is closing is a team sport. And so when we have a candidate, particularly a high profile, high caliber, there are often daily standups around that candidate about what we’re doing to get them to yes.… But someone was sharing, we were working on a candidate in New York and we learned that through getting to know them that they went to a performing arts high school and were a classically trained violinist and was just very into stringed instruments. So this person found this specialty store in New York City and bought this kind of unique Turkish guitar, almost violin guitar thing. And they’re meeting for a dinner where they got the offer and he presented that gift to them. And so it was a couple hundred dollars maybe.… But it’s the thoughtfulness and the touch and the uniqueness of it and the personal moment of that that I think stands out.
B3. High-growth: Multiple recruiting org roles
Tactic: Sourcers and candidate researchers, rebuilt around artefacts
We created this system we call project-based recruiting, where we map out all of the hardest technical problems across all industries that anyone has ever had to solve. We look at temporality: who are the people who did the 0 to 1? Who’s in charge, quote unquote? Who actually did the work? We talk to as many people as possible, and then we just track it. And you’d be surprised by — the number of people who say yes after the first conversation is pretty low, but the number who say yes after the 20th conversation is surprisingly high
Piotr’s research team is tiny, but unusually effective. They often hire by artifact, searching GitHub repos. One early researcher came straight from a call center. Another was building voice models in his spare time.
Tactic: Poach the poacher
there was a guy who was running a next-generation AI chip for one of the frontier companies, and he’s trying to recruit one of our architects. And we actually kind of did an Uno reverse card and started recruiting the person trying to recruit our guy. Within a week we hired him
B4. Executive hires: Retained recruiter
Tactic: Map and recruit the legend yourself (Etched)
if we were going to wave a magic wand, what would the best possible person in the world look like?… So we mapped all of the different teams that related to all of the different rack-scale products at NVIDIA, and we found three people we thought could fit the bill. We talked to all of them, and two of them had just retired, and one of them was planning to do one more generation for NVIDIA and then retire… Over time, we convinced him to join… And the other two guys ended up investing, by the way.
Tactic: don’t hire the famous exec
I’ve actually tried to hire people with some chips on their shoulders who are very talented, but they have not had their first major hit yet. Chips on shoulders put chips in your pockets… putting someone in the waters and letting them figure out how to swim rather than hiring the most well-known expert…