Yonatan Zunger (zunger) wrote,
Yonatan Zunger
zunger

Non-suck resumés

I've spent a lot of time in the past few years reading resumés and interviewing people. A really surprising fraction of those resumés have been really lousy; either uninformative, or full of meaningless junk, or just plain illegible. I tend to throw those out. On the other hand, good interviews with competent people leave me in a better mood all day... so in the interest of getting more of those in the future, I'm going to post a couple of notes on good versus bad resumés. (OK, to be honest: In the interest of not having to slog through any more piles of really bad resumés. It makes me feel like my brains are going to leak out of my ears.)

  • Format: Stick your name and contact info at the top. Then have major categories going down, sorted in descending order by what you're applying on the strength of. (For instance, if you're a new grad, you probably want to put "education" as the top category. If you have professional experience, or internships, or research experience, that's way up there too. You almost certainly want to put a list of your other job-related skills way at the bottom, and other skills below that. Don't skip the other skills; sometimes there are interesting ones there. Fluency in four languages is actually pretty interesting no matter what job you're applying for, for example.)

  • If you're wondering if something is worth putting on your resume: Semi-relevant job experience can be reduced to a bullet point or two under skills. Completely irrelevant job experience, if there's a lot of it, is also a bullet point.

  • This document is meant to be read by humans. It doesn't need world-class page layout, but it should be reasonably easy-to-read; that means well-spaced text, clear fonts, etc. Avoid things like tables with grid lines unless you're a proficient graphic designer and know how to make them legible; 95% of tables aren't, and they do more harm than good. Also, a resumé is an excellent place to demonstrate that you have full command of the language that you're applying in. If you don't speak like a native, run your text by someone who does.

  • Bullet points versus paragraphs: Either is fine, so long as it's informative. A long list of papers you've published is surprisingly uninformative, although I know some resume readers like it. A few lines explaining what you've been working on for the past couple of years is completely critical; no matter how many paper titles you give, I won't be able to fathom anything useful without that.

  • The person reading your resumé is not necessarily a specialist in your obscure sub-field. Telling me that you implemented the G87 patch to the GRU echoing transformer1 is kind of meaningless unless you give me some hint of what the hell those are.

  • Those objective statements... I don't read them. I don't know anyone who does. I suppose they're traditional, but unless they say "To crush my enemies, see them driven before me, and hear the lamentations of their women," they're pretty much just visual decoration. Do not stress about getting them perfect.

  • People say your resumé needs to fit on one page. Damned if I know why; I'm perfectly able to turn a page. Although if there isn't something on the first page that makes me care, I probably won't bother.

  • Really, unless someone is hiring on the "anyone who's at least marginally competent and/or has a pulse" rule, whoever is reading your resumé is looking for one good thing, not a critical mass of so-so things. So padding with a lot of boring stuff doesn't really help; one good explanation paragraph of something cool - even if it's nontraditional, even if it's just some wacky project you've been doing in your spare time - can.

1 This is gibberish, not CS, in case you were wondering. At least, I think it's just gibberish.

Here's a sample of a decent resume, with comments interspersed.


Joe R. Hominid
100 W. Aardvark Dr., Buggersville, CA 94043
tel: 1-415-555-1212
jrhomind@cs.buggersville.edu
http://www.joerandomhominid.net/


ObjectiveA research-oriented position with a practical bent
Education
Ph.D. in Computer Science, University of West Buggersville
Thesis Title: Sustainable Fishing with Bayesian Regression
Advisor: Prof. Moshe Zuchmir		2001-2006 (Expected)
M.S. in Electrical Engineering, University of West Buggersville (GPA: 3.8/4.0)2000-2001
B.S. in Electrical and Electronics Engineering, Southern Turkmenistan Glorious University (GPA: 3.7/4.0)1992-1996
Tells me: This guy is a fresh Ph.D., applying on the strength of that. I've never heard of Prof. Zuchmir, but that may just mean that I don't know him. Maybe one of my colleagues does. There's a time gap between his BS and his MS - I'll check the rest of the doc to see what he was doing then.
Professional Experience
2002-Present: Research Assistant, University of West Buggersville, CA, USA
In partnership with the Buggersville Fisheries Institute, developed a system for continuously monitoring the nearby water ecosystem and directing fishing vessels to optimize long-term sustainability of the fishery with minimal economic impact. One team developed a cheap sensor that monitored eight environmental variables; we deployed over 500 of these. I worked on developing computer models based on Bayesian networks to combine monitoring information with model fishing schedules to predict the long-term health of the region. This included the theoretical development of the model, implementing a working version in C++, and developing and maintaining the system as we used it in real life. Fishery yields increased 15% during the time of our project, versus a 10% decrease for neighboring fisheries.
This is a great paragraph, and I'd pull him in for an interview on the strength of it alone. Why is it good? It shows that the person knows what the overall project is and why it's important (a shocking number of people seem able to work on a small piece and have no idea why they're doing so), it explains what the individual's actual role was (I'm not so interested in a super-duper project if the candidate's main contribution was to sanitize the telephones), and it shows evidence of several important skills: the ability to do the theory work underlying a computer model, the ability to implement a real system and keep it running (don't underestimate that! There are a lot more people with skills in theory than the ability to build a real system. If I'm hiring an engineer, rather than a pure theorist, that's actually the critical bit), the ability to work as a significant part of a larger project, and it's a strong hint that the person must have some common sense.
2001-2005: Teaching Assistant, University of West Buggersville, CA, USA
Teaching assistant for undergraduate data structures, undergraduate algorithms, and the graduate machine learning intro. Duties included teaching sections, developing problem sets, and grading papers.
Also good to know. Less critical than the research experience, so the text is shorter.
1996-2000: Enlisted Sailor, US Navy
Served in the US Navy as fire control aboard a number of surface ships. Honorably discharged at the rank of Petty Officer 3rd class (E-4).
Even though this isn't technically relevant, it tells me what was happening during the four-year gap, and any serious job you held for that long is interesting. For the purpose of this resumé it's mostly flavor, on a par with the "other skills," but it gives me a sense that this candidate is used to discipline and professionalism -- potential strong points. E-4 is a bit of jargon, but "PO3rd class" explains it partially and there's no other good way to encode that information.
Etc. etc.
PublicationsA big list goes here, I don't feel like making up a bunch of entries
I'm not going to read this in detail, but if he has a range of publications that's interesting to know, especially if they're not all with the exact same list of authors. (That would just tell me that the principal investigator of the group is the kind of guy who publishes everything) If he has conference papers, invited talks, etc., those should get top billing and maybe their own subsection.
Technical Skills
  • Fluent in C++, C, Python, x86 assembler
  • Experience with machine learning techniques, especially Bayesian networks and SVM's
  • Was sysadmin for a 20-machine Linux cluster for the ecology department, 2001-2005.
Sysadminning is one of those "partially relevant" jobs. It's evidence of some real-world experience, which is good, but for a research-type position it doesn't get top billing.
Other Skills
  • Fluent in English, Mandarin Chinese, Classical Greek
  • Volunteer firefighter, Buggersville Fire Department, 2000 - Present
Massively relevant? No. Shows that he's got a range of interests, can probably think about things other than the narrowest details of his field, and is generally kind of neat? Yes.

Disclaimer: This is not the official position of my employer, or anyone else except me. I offer no guarantee whatsoever that if you follow this, you will get interviewed, hired, even noticed by a potential or current employer, or even not be shot at by them. Nor is any warranty, express or inferred, offered in conjunction with the reliability or usefulness of this advice.
Tags: computer science
  • Post a new comment

    Error

    switch
    default userpic

    Your reply will be screened

    Your IP address will be recorded 

    Help
  • 17 comments
...go 1 level up... Expand

zunger

November 18 2006, 00:06:06 UTC 9 years ago

And it would definitely have a higher probability of getting read. :)

kamileon

November 18 2006, 02:09:28 UTC 9 years ago

And if the company calls, you know you are guaranteed to have a hiring manager with a sense of humor, which is definitely nothing to sniff at. :)