Hacker News Reader

Comment by tjpaudio | Hacker News Reader

Definitely. I have a high-paying data science job at a major corporation with only a 4 year degree in economics and got it after 2 years in the field working analyst jobs. I took stats classes with my free electives, got database skills from my first job and taught myself how to code along the way. Many of my peers have advanced degrees and I can hang no problem. I will say though, that the data science boot camps are not the answer. The applicants we receive from them are generally weak and we have never hired any of them.

View on HN

you said none of the bootcamp people that applied were hired. out of those bootcamp people, how many had bachelors, and what was the distribution of their degrees? i.e. how many had degrees in CS, Math, Physics, Biology, etc. Alternatively, out of the math degrees that applied (regardless of bootcamp), how many were hired?

Of the bootcamp people, everyone has had at least a bachelors, maybe weighted towards less stem but that's certainly not the rule. I don't think I have a good numbers backed answer for you, but I also don't think prior academics is the issue. It's mostly peoples ability to problem solve. We have no shortage of applicants but it still takes us forever to find people that can do the job.

You mentioned people's "ability to problem solve", and that it "takes forever to find people that can do the job". This is very interesting. How do you measure "problem solving ability" in your applicant pool when all you have are resumes? Do you simply look for points on the resume which reflect positive improvement? Do you have more than just a resume to base your judgement off of?

E.g. consider these two similar points on a resume: "improved loading times tenfold by using compound indexes and denormalization" would reflect an improvement from a problematic situation to a better situation. Whereas "designed schema and implemented indexed database" does not explicitly name a problem and its solution.

I hope my point makes sense, and I am wondering if little differences in the wording make a big impact in how you perceive someone's ability to identify and solve problems.

We give a data set and ask some open ended questions. Out of 800 applicants, maybe 50 complete the exercise, and then maybe less than 10 we'll want to talk to based on what they sent. I spent maybe 5 hours on the dataset when I was interviewing for the position myself.

Wow, only 6% complete the task? I am testing my luck here, but do you have any sample/link to a data-set or questions you have asked before, or maybe a link to something similar to what you have asked?

The dataset is private but the general gist is its like 100,000 rows with conversion and revenue data with time and 2 other dimensions. There are some outliers (clearly errors) in the data and strong seasonality differences in some dimensions, and data sparsity in others. The applicant isn't told about these, but we look for them to find it in their analysis and control for it in a predictive model they are asked to build based on the data.

Reply on news.ycombinator.com