by tdhopper on 3/6/17, 2:51 PM with 165 comments
by SatvikBeri on 3/6/17, 7:50 PM
1. Break down "data science" into several different roles–in our case, Analyst (business-oriented), Scientist (stats-heavy), Engineer (software-heavy). Turns out that what we mostly want are Engineers-Analysts, so our process screens heavily for those.
2. Figure out which types of people can be trained to be good at those roles, given the team's current skillset. I opted to look primarily for people with strong analysis skills and some engineering.
3. Design interview tasks/questions that screen for those abilities. In my case, the main thing I did was make sure that the interviews depended very little on pre-existing knowledge, and a lot on resourcefulness/creativity/etc. E.g. the (2-hour) takehome is explicitly designed to be heavily googleable.
4. Develop phone screens that are very good at filtering people quickly, so that we don't waste candidates' time. By the time someone gets to an onsite interview on our team there's something like a 50% chance they'll get an offer.
On the candidate side, when I'm applying I try to figure out first and foremost what a company means by "data scientist", usually by networking & talking to someone who already works there. This filters out maybe 90% of jobs with that title, and then I put more serious effort into the rest.
by kafkaesq on 3/6/17, 8:05 PM
BTW, as to some of that "feedback":
Honestly, I think the way you communicated your thought process and results was confusing for some people in the room.
"Okay, pal - let's put your people up in a room full of strangers (some of whom show through their body language and/or constant phone-checking that they pretty much don't really want to be there in the first place), make them answer some made-up questions (which may or may not be coherently presented; or even particularly relevant to the job description) -- combined with the background stress of being unemployed, and/or stuck in job they absolutely HATE, and can barely stand another day of -- and see how they do."
Quite honestly given your questions [about vacation policy] and the fact that you are considering other options, [we] may not be the best choice for you.
Great -- so they're basically accusing you of being a slacker (not really into the work, only interested in what's in it for you, etc). Which is quite a presumptuous attitude to take in response to a perfectly reasonable question about the value proposition they're asking you to consider (a question for which it might be in better form to wait until the later stags of the negotiation process to bring up... but that's a very minor style point that you definitely shouldn't be dinged for).
"Quite honestly" that attitude sucks. And you don't need to feel bad about being "rejected" by people like that.
by minimaxir on 3/6/17, 7:30 PM
Internal recruiters have hinted that my Software QA Engineer background + no CS degree implies I have no technical skill.
by _flbt on 3/6/17, 8:06 PM
Once I got that offer and 'data scientist' was listed as a position on my resume, I've had at least one or two recruiters reach out to me each week. Not just your 'bulk tech recruiter', but individual hiring managers/team leads from companies who had no interest in me prior to the title change.
Hell, I had a manager I had already interviewed with from another company (and got rejected) reach out to me TWICE to come back in. To be honest, I'm not entirely sure he remembered that I had interviewed there only a month earlier.
All after I had the title change on my resume. What I'm getting at, is for me, the title change really opened up doors. I figure 'data science' is such a huge buzzword now, recruiters look more for that buzz word than they do for the actual content of what you've done on your resume. I'm sure it will die down at some point.
This may be more prominent outside of silicon valley where I am, and (in my opinion without any facts to back it up), where I feel more weight is given for the title than the substance.
by graphememes on 3/7/17, 6:54 AM
I've had the following:
- 106 Rejections
- 39 Offers
- 24 Refused
- 15 Contemplated
- 5 Taken
- Software Engineer (55rj, 8o, 6r)
- Product Manager (5rj, 6o, 5r)
- Senior Software Engineer (31rj, 23o, 22r)
- Product Lead (0rj, 1o, 1r)
- Senior Product Manager (9rj, 3o, 2r)
- VP of E (2rj, 1o, 1r)
- CTO (0rj, 1o, 0r)
Things to note:
- People conflate Agile & Scrum a lot.
- Make it easy for people in the room to know what you personally accomplished during your career.
- Make an impression, don't overdo it however.
- Make your descriptions easy to understand.
- Don't just list tech stacks on your resume, list achievements.
- Stay consistent.
- Ask questions.
Regardless how impressive your Github & Stackoverflow accounts might be, you will still be asked to do a stupid code challenge. It irks me. I understand the reasoning for it when you don't have a viable pool of information on the individual, but when you do... It's just disrespectful.
Few places I have been rejected by:
Salesforce, Atlassian, Trello, Twitch, Segment, AirBnB, Twitter, OKCupid, LinkedIn, Shippo, and many many more.
Amazon, VMWare, Oracle, Google (twice), Apple (twice), Venmo, Auth0, Discord, Youtube, Hitbox, Steam, Pusher, Cloudflareby strictnein on 3/6/17, 7:55 PM
I had a very similar experience. Job offer was basically on the table and then they balked because I mentioned that I had another offer (at a larger company, which they seemed shocked/annoyed with) and I had a question about parking at their new offices. The current offices had free parking, so I had simply asked if the new ones would too. The response was very cold (must have a been the source of internal strife, I guess?). An hour later I got a call saying they were no longer interested.
by motivic on 3/6/17, 7:09 PM
To share my story, I also had a difficult time transitioning into a data scientist role after leaving academia (pure mathematics), and I always thought the root cause was my lack of experience and competency. So instead of keeping on applying, I spent over a year just to sharpen up my skills. It paid off in the end.
How can one develop his/her skills and cultivate expertise if one is job-shopping all the time (possibly aimlessly)?
by bsg75 on 3/6/17, 8:19 PM
Dodged a bullet on that one:
- PTO is a touchy subject?
- They are looking at more than one candidate, why would any candidate limit themselves to one potential employer?
by preordained on 3/7/17, 11:59 AM
I have a hard time understanding. At least for what I've seen/done, it would take about a year of experience to be of any real value. I'd say you start seeing real dividends from an employee near the three year mark. I think the primary exception would be where you have a big gaping hole in an organization...like building a data science program or something from the ground up. But if you've got software and customers long long past the 1.0 stage to support, and someone is going to (someday) understand/contribute to the core? Think about Google's monolith or the Linux kernel...sure most software isn't that mature/grand, but there are many projects out there that are closer to that than greenfield. And it's not just the code, it's the developed relationships/rhythm with coworkers and customers.
Maybe I've explained it to myself, I don't know. The difference may be startup companies or projects versus mature ones. At least in the latter case, it seems to me that if a company retains an engineer for less than 2-3 years, they've almost certainly lost on that investment.
by alain94040 on 3/6/17, 10:34 PM
That's actually the only correct answer. Having been on both sides of the table for many years, I can pretty much guarantee that whatever reason the candidate is given is nowhere close to the actual reasons.
There may not be any specific reason why we didn't pick you, but we'll give you a tiny sample anyway. So you think that's the reason - it's not.
In other cases, we have a strong reason not to pick you, but it's embarrassing so we feed you a bogus reason instead.
And in more cases than I'd like, there is no problem on your side, we are having internal issues that we can't reveal anyway. Any reason you hear from us in that case is completely irrelevant.
By the way, when you evaluate people in interview, you really need to figure out a vector: where they are today in terms of knowledge and experience, but also in what direction they are going (fast learner or not, high potential, etc.). Which is why sometimes you can hire someone who has less relevant experience, but you think they'll learn fast and are very smart, and sometimes you are looking for the perfect match to the current position, but don't care too much whether they can pick up new stuff or not.
by CSDude on 3/6/17, 7:41 PM
by pascalxus on 3/6/17, 6:36 PM
by CalChris on 3/6/17, 7:05 PM
For startups, this transcends data science. It might be the one time that week they focus on that need.
Networking is still king.
Exactly and this also argues against wanting to get hired to work remotely.
by eanzenberg on 3/6/17, 7:15 PM
- Signal is still quite low among noise, even with long multiple interviews, take-home homework, coding challenges, etc. Most relevant data is still hidden and takes months-years to come out.
- Companies seek to minimize false-positives much more than minimizing true-negatives.
- It's a numbers game from both ends because the probabilities are low, due to above 2 points.
by donovanm on 3/6/17, 7:39 PM
by autokad on 3/7/17, 12:43 PM
Does anyone have any suggestions? Here is my linkedin account: https://www.linkedin.com/in/karl-dailey-02557b65
by rubayeet on 3/6/17, 6:12 PM
My takeaway from it.
by ENGNR on 3/7/17, 12:50 AM
by segmondy on 3/7/17, 2:47 AM
There is a ramp up time for new hire, which could be a couple of months. So durations of a year or less doesn't look good. I personally like to see minimum of 2 years for each job. Of course, too long can be a concern too unless they really grew in their role.
I agree with his conclusion, Network is king, but I also believe in listening to the universe. If everyone is "wrong", maybe you are the problem. If lots of people don't wish to hire you after you have solid experience in the industry, something is wrong with you and you are being stubborn by refusing to recognize it and fix it.
by mattfrommars on 3/9/17, 6:36 AM
My future really rests upon this. I have a degree in mechanical engineering with irrelevant experience. My only regret is not doing bachelors in computer science.
I wouldn't lie if all the post got me little tensed and made me thinking, "what am I doing with my time". About to finish introduction to Computer Science using Python on Edx.
by carterehsmith on 3/7/17, 2:06 AM
That said, what is this?
a) "I cannot do X" b) "But here is some advice on how to do X"
I mean... why would anyone listen to the advice that is proven to lead to failure? Serious question.
by LeanderK on 3/7/17, 5:28 PM
I know switching jobs is common here, but i would think that sticking at least 1 to 2 for a job would be normal (assuming it works out).
by megablast on 3/6/17, 11:03 PM
Bit of a silly article. You could say this about anything. And I don't see what the problem is with never being contacted, if someone doesn't want me for whatever reason, I don't really want to hear from them again.
by anotheryou on 3/6/17, 8:20 PM
by edblarney on 3/7/17, 8:43 AM
You wouldn't want to work there.
" but the team has decided to keep looking for someone who might have more direct neural net experience."
Fair enough - but this has to be slim pickins. How many AI jobs are there out there? Realistically, very few.
by draw_down on 3/6/17, 7:58 PM
We got confused, and then we stopped. The end.
by gillianlish on 3/6/17, 11:23 PM
apply for 100 jobs, get zero response, and zero reason why.
be told illogical things.
be told out right lies.
welcome to capitalism. welcome to the workplace that is run entirely by data. (or by people who only care about data). welcome to the future of humanity.