So, what IS a Bootcamp?

Hay una versión en español de este post, aquí.

This is my first post on a series of many to talk mostly about Data Science. I decided to start it to talk about my experience on Metis Data Science Bootcamp and later on, my exprience, projects and insights as a Data Scientist!

About me

To get a better context on what this blog will be about, here’s a little personal context:

I’m Colombian, I live in Bogota, and have a Computer Science degree there. I have worked for several (10) years in Data Warehousing and Big Data Frameworks. I’m passionate about coding, programming and algorithms and a couple of years ago fell in love with all these Machine Learning, Linear Regression Models, Clustering, Predicting… What I like to call “Playing with the data” and some people call “Data Science”. So I decided to study more about it. They say that a Data Scientist is either a statitian who’s very good at programming, or a programer who’s very good at stats. I’ll try to be the latter.

Why a Bootcamp?

So I got the programming skills covered. But I’m rusty on all the statistics, math and probability you need to know for these stuff. There are a lot of online courses that you could take, most of them free and high quality. That works for many people, but for me it was a little hard to get the free time between my job and my personal time. I knew that, in my case, it had to be on site. Now I thought about a masters degree, or a PhD… and there are many of them out there! Most of them one or two years long. But lets face it: This world, and this area of expertise changes very fast. What you learn now will be useless in, maybe a year (?), so it’s important to learn the basics and what’s underneath, but not spend too much time on the technology that, either way, we have to keep studying and updating ourselves. So that’s what a Bootcamp is. A new methodology that, in short time (12 weeks, full time), will cover all the topics you need, in one or two technologies, that will require a lot of time and self study. Seems like an assisted information injection.

And how does it work?

To apply for a Bootcamp, you generally have to take an exam to assess your level in: Programming, Linear Agebra, Calculus, Probalility and Statistics. Your classmates will be people from many different areas (In my cohort I met, from physicists to designers! and, curiously, very few programmers like me). The idea of the exam is for all of us to arrive with the same level in all those topics (But not in “level: zero”). Accordig to the results of that exam they tell you if you pass or you have to commit to study deper into one or two of them. My recommendation is always pay ttention for any concept you don’t understand, do your prework before beginnig the Bootcamp, and everyday make questions, look for lectures or videos on any concept you don’t quite understand.

A Bootcamp is based on projects (every two or three weeks!). You’ll feel that time’s not enough but you’ll be surprised of of the amount of think you’ll get done and learned in so little time, and that’s the idea. The theme of each project is completely your choice, but you have to cover different topics everytime (Regression, Classification, Supervised and Unsupervised learning, NLP, Neural Networks, …) So projects are as hard as you make them, as much as you push yourself. You also have to fulfill other assignments or homework very week, create a blog (and get the habit of writting blog posts frequently: that’s where this blog came from!), research and presentations related to AI or tech… It seems lika a lot, but it’s well structured; make sure to accomplish everything because all activities will help you undesrtand and develop your projects that are the most important thing!

The instructors are experienced Data Scientists, some with even a masters or a PhD. There are generally two principal instructors per cohort and the interesting part is that they’ll give you two different points of view to attack a problem. They will not give you the straight answer to a question, but guide you to find it yourself (This can be frustating at the moment, but you’ll see that you’ll learn a lot more). They know a lot of resources, books, blogs, people, where you can go and learn more. But most importantly, they’ll help you conquer your project, no matter how crazy it is, guiding you, giving you ideas, and showing methodologies that you can only get with practice and experience. Sessions are organized as follows: Pair Programming exercises every morning (short and quick) to wake up your brain, main lectures given y the instructors, and project time (like free time to work in your own project, but the instructors will always be there to answer your questions).

Career Support?

Most Bootcamps say they’ll give you orientation to face your career to this new role: Data Scientist. There are a lot of practical workshops like: upgrade yout CV, your LinkedIn profile, develop your networking and communication skills, mock interviews and workshops, talks by profesional Data Scientists from other companies about their experience where you can ask them anything. After Bootcamp, people from career support will stay in contact with you for advice, questions, mock interviews or anything you need.

In my case, as an international student with plans to return to my country, I didn’t make use of other resources they offered like: Visit to recruiting companies, salary negotiation workshop, networking events with partners; but I heard that these were also very helpful for my classmates.

Conclusion

Bootcamp experience is amazing and very productive. You get the tools and knowledge needed to start working as a Data Scientist and the habits to keep studying and growing. Me, for example, after just one month after returning home, I was working in a tech consultant company, in the Data and Analytics area, mi new role: Data Scientist.

The methodology gives you experienced instructors, but it’s with discipline, self-learning, reviews, research, discussions and collaboration with your classmates that you get to understand thing completely and get deep into the concepts.

Prices in USA are anit high in my personal opinion, but it’s totally worth it. I’ve already heard of the same methodology applied to Data Science, Cloud and Big Data in some countries in Latinamerica!

If you’re an international student, don’t expect a job offering. But hwat you can expect is returning home after 12 weeks, with all the concepts, tools, knowledge and, most importantly, learning habits that most people take one year or more to learn and succeed.