The Kleiner Perkins Fellows program is a fellowship for all of Engineering, Product and Design which gives students the opportunity to intern at a portfolio of early and late stage startups while being able to engage and learn from a broader community of other students’ startup experiences. The fellowship is geared towards those not just with solid technical skills, but exemplary leadership and strong evidence of a self-directed entrepreneurial mindset. As a KP Engineering Fellow in 2014 myself, I currently work at an early-stage Kleiner-funded Series B startup called Glean where I was on the founding team. Many ex-fellows have gone on to do some amazing things, including founding large, growing startups like Figma, OpenSea and Persona, holding senior positions at Doordash, Duolingo and Quip and going on to do PhDs at Stanford, MIT (HCI, Bio, CS) and Harvard, and many, many more.

A small selection of some past Kleiner Perkins Fellows.

Statistically, the KP Fellows program now has 2500+ applicants from 215+ universities primarily in the US and Canada and last year they had a total of 102 fellows, 71 in Engineering, 13 in Design and 18 in Product. That’s a ~4% acceptance rate. Of these applicants, 40% are female and 60% are upperclassmen (or Masters students). Since it’s inception in 2012, Kleiner has produced 732 fellows over 10 years.

I’ve been on the selection committee for the Kleiner Perkins Fellowship for several years now and have conducted 100+ technical and behavioral interviews across the fellowship and my experience at Glean, Google and Facebook. Prior to the semi-final stage of the KP Fellowship, I scan through 1000s of resumes for this and have the dubious distinction of having created a semi-viral resume template of my own a long time ago. When I was an applicant, I found applying to jobs and programs like this to be a stressful, byzantine process with limited insight into how it works on the other side. As a judge now, I thought it would be valuable to point out certain things that applicants to technology companies should keep in mind when they approach this process.

How the Process Works

Kleiner Perkins Fellows applications are rolling and decisions are made typically around April-June. Candidates submit 2 personal statements, a resume, domain challenges based on the track they’re applying for, and personal links (i.e. medium, youtube, twitter, github, etc.). During this review, a sizable percentage of candidates are filtered out easily because:

  • They didn’t complete the domain challenge
  • They didn’t answer the essay prompts
  • Their field of specialization doesn’t particularly lend itself to what the role requires

Applications are reviewed by the selection committee which comprises KP Fellow alumni who look for qualities such as: technical characteristics, leadership, entrepreneurship and communication skills. Reviewers are allowed several “Bets” on candidates who may not fit the typical criteria but have a phenomenal story to tell. As an applicant, I would think that this is anti-meritocratic behavior, but now that I’m on the other side, some of these stories are truly inspiring and if other applicants had the chance to read them, they’d agree. Then, by around March, you move on to interviews both with the Fellows in the semi-finalist round and with KP portfolio companies you’re interested in. On occasion, candidates have already accepted an offer from a port-co and only need to pass the semi-finalist interview. After these interviews, final decisions are made and released to the public.

Common Resume Mistakes

Before I list out specifics, I ask that you take a couple minutes to empathize with the job of the reviewer. They’re spending valuable time on a weekday or weekend outside their hectic daily schedule to voluntarily review a boat load of applications. This process isn’t perfect. We know that humans recognize patterns and biases may come into play. That said, each member of the selection committee is educated thoroughly on understanding these biases. Every candidate will also have their application reviewed ~3-4 times from various members of the selection committee. Remember that while there are many things you can do to increase your chances, sometimes you just have a bad reviewer having a bad day. If the entire process were to be repeated with the same applicant pool, I imagine the results could turn out differently.

For the purpose of this post, I’ll be writing specifically about some common resume mistakes for applicants to the Engineering Fellowship.

  • Don’t expect a reviewer to spend more than 5 minutes on your resume. We know that resumes are a pretty poor representation of a person’s potential which is why they ask for two personal statements, a domain challenge, and personal links. From what I’ve seen from reviewing a large volume of applications, I know that there are times when I anchor on certain patterns to move quickly. We don’t have hard and fast rules. Sometimes, I skip in 30 seconds (of course, we require more time to review the candidate’s personal statements). Sometimes, if I see an appealing format, I click into more links to learn more and visit the candidate’s website, LinkedIn, Github and more. Contrary to urban legend, there’s never a need to visit a candidate’s personal social media.


  • We look for more than college, GPA and brand-name internships. Our goal is to build a diverse and entrepreneurial cohort. We know that things like brand name internships, GPA, Ivy League are not a good measures of success and have held less weight over the years as we try to find the next generation of leaders. However, when reviewers have to go through a lot of resumes, humans tend to look for patterns which they think have high signal to noise ratio. Given how hard it is to assess a listed programming language skill, a listed course, Github projects and the actual work you have done at a past company, sometimes reviewers may anchor on these basics. We look beyond colleges, internships, grades, and have a focus on finding unobvious candidates as talent can be found anywhere but opportunity is not. We ensure applications are received 3-4 times to counter this bias.


  • If your experiences are mixed, make sure it tells a cohesive story. We love to see a broad and diverse range of academic and professional interests. We see a lot of double majors (the most common one today is Computer Science and Economics), a variety of internships, and a broad range of activities. While we love people with well-rounded interests, it’s also important that your application tells a cohesive story. Sometimes, too varied an experience can prevent you from standing out at anything in particular.


  • Repeatability is underutilized. People like to put people in buckets. When reviewing a resume, if I see multiple projects that involve donations to charity, I believe that this person is genuinely interested in that. If I see multiple web3 projects, I believe this person genuinely cares about web3. Use repeatability in your resume across the different sections to tell a coherent story.


  • Using class projects as “Projects” may seem like a hack. One of the coolest things about Computer Science is that with a laptop and internet, you have the resources to pretty much build anything you want. It’s a blank canvas to be creative and useful. Doing a cool or nifty project on your own, especially one that solves a critical problem for yourself and others, is a strongly positive signal. That said, it’s important to distinguish between projects that you did off of your own accord and projects you did for class. Using class projects as “projects” seems like a hack. Often, these may have a lot of boilerplate code, don’t typically demonstrate that you sought out an interesting problem, and don’t exhibit creativity. How do we know they’re class projects? Because your other friends from that college use eerily familiar words to describe it. That said, not all class projects are bad and occasionally classes allow you to make your own project. I’d just appreciate it if class projects were clearly labeled as such.


  • Be thoughtful when using the term Co-Founder. It can be very confusing to me when applicants call themselves a Co-Founder of a project. It also doesn’t make much sense to see a current, full-time startup co-founder is currently applying to internships. Of course, this doesn’t apply to clubs, podcasts, non-profits, and more! Use your judgment wisely. Based on what I’ve seen in the industry, here are some conditions that make you eligible to be called Co-Founder of a startup:
    • Solves a problem and has relevant traction to support it
    • Whatever you do makes money, or has a path to it
    • You raised venture capital
    • You have at least 1 person on payroll


  • Controversial: I like seeing a glimmer of personality. Are you an athlete? Did you do something unique? Is there something outside of work you’re good at? What sets you apart? What’s your main thing other than the gig you’re applying for? As someone who works at a startup, it’s very important to work with people who are competent but also cool and interesting. I like seeing things like these to validate that you’re not just buried behind a computer your whole life (and even so, there are so many awesome niches to explore digitally too).


  • Personally, I like seeing published research in good journals. There may be other reviewers who don’t care for it, but to me it signals a high level of dedication to a problem that is vetted. Others are of the school of thought that good research is not a signal for good engineering.


  • Personally, I like seeing Olympiads and competitions. Candidates who participated and excelled in USAMO, USACO, or AIME, IMO, IOI, or their respective country’s national or international olympiad in Math, Physics, Computer Science, Chemistry or Linguistics usually possess deep problem solving ability. Olympiads that are not stepping stones to the official international olympiads usually aren’t as credible.


  • Resume formats kinda matter. A bad resume, usually one which looks like they printed a Word doc to a pdf can be hard to review. When applying to technology companies, I don’t think anyone minds colors on your resumes. If you’re a designer, and your resume design is poor, I might think twice. Don’t cram too much text in your resume. A good resume won’t make your entire application, but it does send a soft signal that this person went out of their way to create this. It leaves a subtle mark of competence in the reviewer’s mind no matter how unbiased they try to be. Smart people can still have poorly formatted resumes but it handicaps their odds unnecessarily.


  • Links are necessary, but might not get clicked as often as you think. I know reviewers who meticulously click links to dig deeper into an interesting project or website. Personally, I wish I clicked them more, but the cost of context switching, inability to click on links while using Preview to view a resume, the varying formats of different personal webpages, and the sheer time it ends up taking makes it something I do only for certain candidates that pique my interest.


  • Let me tell you a long list of project clichés. Not all of them are bad. It’s really difficult to evaluate the competency of projects, so I wish candidates would brag more. Attach a link. Attach a Github repo. How many downloads did it get? Views? Did it feature in an article or HackerNews? How many users does it have? Does it make money? Was it #1 on the App Store? What languages or frameworks did you use to make this? Personally, I like projects that seem genuine and not resume-padding - does it solve a minor but real user need you probably faced? Not the class of “hammer looking for a nail” projects where you use a fancy technology for no practical purpose. Here’s an extended list:

    • Games and AI solvers for games: Scrabble, Chess, Pokemon, Chinese Chess, Connect4, Sudoku, Mazes, 2048, Wordle, Poker, Othello, Blackjack, Checkers, Breakout, Tetris - Impressive if not for a class, depending on the game. I see a lot of “minimax with alpha-beta pruning”. Waiting for someone to make one for Catan!
    • Course plan visualizer for their college - I find this very impressive, especially if they have a working link and other students actively use it.
    • “I trained a model for X to do Y” - Very hard to evaluate, and I assume you ran <100 lines of code in a Python notebook most of the time. Variations: Image colorizer, object detection, text summarizer, predicting diseases, OCR, speech to text
    • “Co-Founder at personal project” - very confusing to me.
    • Sign language to text - An eerily common theme. I am fairly neutral to this project, because like ML projects, the quality makes or breaks it, and I can’t judge the quality.
    • An iOS / Android app - I like it, but I cannot tell if it’s good unless you tell me how many downloads it has.
    • Facebook / Twitter / Messenger clones - Mixed. These are common college class projects.
    • Text summarizer - Too hard to evaluate. Seems like you plugged-and-played an ML model.
    • “Covid tracker” - Impossible to evaluate, especially without a website to go to.
    • Stock market predictor - Another classic cliche. Also impossible to evaluate or trust any results.
    • ChatBots - Hard to tell if essentially a few if-statements, or more robust than that.
    • Speed Reader - Interesting how common this is! I think this is quite cool albeit simple because it’s usually indicative that this person hacks on stuff for themselves
    • “Mental health app that provides resources for XYZ” - It’s unclear to me what this means in practice
    • Haiku generator / Text generator - I actually like these because it indicates you are passionate about coding for fun
    • Web3 projects - DAOs, NFTs, Trading bots - I’m not sure how to evaluate these without seeing some impact numbers.


  • “Objective”, address, citizenship status, and other minor things etc

    • Please don’t have your resume spill over to two pages unless you have so much published research that you require this.
    • Objective / Summary - definitely not necessary, but if you have a very specific interest or location restriction, I could see it being useful. Don’t add if you’re going to say something generic like “high-impact leadership role which combines business and engineering”. Please keep it short - one line, preferably.
    • Address - These days, absolutely unnecessary.
    • Citizenship status - I imagine this information is useful for a certain subset of employers, although typically I’ve not seen it make a difference for Software Engineering roles.
    • Github link - Please have one.
    • Website - I think it helps, and you should have one if you’re a Computer Science major, even if it’s a static github.io. Using Squarespace as a CS major is a poor look.
    • Coursework - I understand it’s pretty typical to list, but if I have to be honest, I don’t think I have ever read this section.
    • Please don’t list Excel and Word as your Skills

Making Your Resume Stand Out

Listing a long list of “don’t”s without a few “dos” is probably not the most helpful. While most application reviewers agree on what they don’t like to see, I hesitate more to make sweeping statements with regards to “dos” because reviewers can vary significantly in what they do like to see. If I had to chime in with a single data point, I’d advise:

  • Formatting
    • Don’t shy away from using a colorful, unique resume format.
    • Keep it in one page unless you have plentiful published research.
    • Don’t use so many colors that it hurts the eyes.
    • Avoid skill pie charts or percentage skill bar graphs. They’re meaningless.
    • Resumes formatted with LaTeX instead of Word show me you put in an effort.
    • Ensure the PDF links to all the relevant places it needs to - your website, your github, your social links, etc. Make sure none of those links are broken.
  • Education
    • Generic coursework is usually unhelpful and consumes unnecessary space. List only unique, challenging ones.
    • When you list a scholarship or an award, it would be nice to have some context on how many people get that award and what it’s for.
    • I’d encourage you to list your GPA, and not forget to list the correct denominator. List your major GPA if it’s significantly better than your overall.
    • If you have an accurate gauge or award that reflects your percentile in class, list it.
  • Projects
    • A project only feels real if you link to a demo and a Github.
    • I love to see people brag about their projects. Tell me if it was #1 on the App Store, featured on Hacker News or Reddit, has many views, or has 100s of stars on Github. I want to know if people loved it! This makes your resume stand out more than just an internship at Google.
  • Work
    • If your internship touched a public consumer product, I’d like to know where.
    • If you did multiple internships at the same time, an explanation would be nice.
    • If you have a long-standing personal “job” over many years, please list it as part-time.
    • Mention if you got a return offer!
    • Don’t exaggerate. As full time employees, we know what projects interns typically get - don’t use your team’s entire mandate as a proxy for the work you did.
  • Other
    • There is no need to list Word and Excel as skills. I’m pretty on the fence about CSS and HTML too.
    • Personally, I like seeing cool interests. List them! If you’re really good at something like chess or rock climbing, throw in a rating or a climbing grade.
    • Not many people do this, but if you have a bunch of public media features for things you’ve done, be it Youtube, a blog post (hopefully not your own), a news article, or more, list it with a link!

Application Statistics

These top of funnel application numbers are for the Engineering Fellowship (interns).

Distribution of Gender

  • Male: 61%
  • Female: 39%

Distribution of Year

  • 2021 (Masters): <1%
  • 2022 (Senior): 8%
  • 2023 (Junior): 46%
  • 2024 (Sophomore): 30%
  • 2025 (Freshman): 15%

Distribution of Degrees

  • Computer Science: 68%
  • Engineering: 15%
  • Math: 3.5%
  • Other: 13.4%

Distribution of Bachelors Programs (engineering only)

Of a total 785 reported after the first round of filtering, here are the 10 colleges with the most candidates:

  • University of California, Berkeley: 11%
  • Stanford University: 6.4%
  • University of Southern California: 5.7%
  • University of Pennsylvania: 5.1%
  • Georgia Institute of Technology: 4.7%
  • University of Texas, Austin: 4.6%
  • University of Waterloo: 4%
  • Columbia University: 3.7%
  • Harvard University: 3.4%
  • Washington University in St. Louis: 3%

It’s unsurprising to me to see some of the more popular Computer Science universities at the top. Many of them have very large CS programs which bias the data. Even knowing about the existence of such fellowships biases towards such universities. The trends are promising, and the proportion of candidates from, for example, smaller liberal arts colleges and historically-black colleges and universities have grown significantly in the past years! As a reviewer, I actually wish I’d see more applicants from smaller CS programs and lesser-known universities and I’d like to personally encourage students from colleges not on this list to apply to the Kleiner Fellows Program in the years to come.

I love hearing feedback! If you don't like something, let me know in the comments and feel free to reach out to me. If you did, you can share it with your followers in one click or follow me on Twitter!