About a year ago I wrote a review of The Data Incubator (updated review is here). I always know when the Data Incubator application season is here because I always get a few people who have found my blog reaching out with questions about the process. I decided to put together a short list of some of the most common questions I get asked.
Should I do the program?
The program has pros and cons. For some people I think the program is great. It provides an overview of a lot of data science topics and practice with concepts. However, it’s probably not as useful for people weak in programming or those that hope it will help them network or get them a job in a specific area/industry. In general, if you’re not accepted as a Fellow (tuition-free), I think it generally isn’t worth it. You can read more about my thoughts on the program in my review of it.
How good is the network of hiring partners?
Much worse than you would expect from the slick marketing. Around the areas the program is located, especially NYC and the Bay Area, there are a fair number of hiring partners. However, the types of industries are pretty limited. Don’t expect to be able to work in gaming or non-profit, those jobs are extremely rare despite being listed alongside industries with way more hiring partners (like finance). If you have strong geographic constraints, don’t expect much from the program’s network.
What is the admission process like?
It’s long and involved, especially compared to other programs like Insight. There are three stages to it. First there is an initial web application that is just a typical application asking for relevant background, education, why you want to do the program, where you could work, etc. If you make it through that stage, you are considered a semi-finalist, where there are some more involved challenges thrown at you (see below for more info). Finally there is the finalist stage, where you go through a quick interview.
What are the semi-finalist challenges like?
The hard part of the semi-finalist challenges are a couple of tough programming questions. You have a limited amount of time to do them (~4 days), with no flexibility about the schedule for doing them. Being very comfortable with Python will help a huge amount here. For a sense of the kinds of questions they like to ask, check out their (blog post on efficient numerical computation)[https://blog.thedataincubator.com/2015/01/a-cs-degree-for-data-science-part-i-efficient-numerical-computation/], they love asking you to make use of the concepts there.
In addition to the programming challenges, you’ll be asked to come up with an idea for your capstone project and submit a short video describing it. They also want a preliminary data visualization, and like for it to be interactive (best is if it’s hosted on Heroku, see their post on getting set up on Heroku).
I’m a finalist, what is the interview like?
The final interview is pretty superficial. It will happen with a group of other candidates. You’ll have 2 minutes (seriously, that’s all) to pitch your project and show a couple of preliminary analyses. The other candidates will have a chance to ask one or two questions. Then it will move on to the next candidate. The interviewers won’t ask any technical questions, but will ask things about where you’re willing to work. For the whole group of 4-6 candidates, the interview will only be 30 minutes.
What should I do for my capstone project?
It’s nice but absolutely not necessary to have a project dealing with data in the industry you would like to work in. Don’t be overly ambitious, make sure you can get the data pretty easily, and just plan to build a predictive model of some kind. For your video and initial data analysis, keep it simple, all you’re doing is showing that your idea is feasible and interesting. When you start working on your project, having a nice presentation/app is usually more important than building a well performing model.
That’s all for now. Feel free to reach out with any additional questions not asked here, I would be happy to add to the list!