Photo by Aral Tasher on Unsplash
We recently hosted a Live Online Ask Me Anything (AMA) session with Metis bootcamp graduate Leon Johnson, a passionate musician, philosophy (AI) enthusiast, and was recently hired as a Data Scientist at Viral Launch in Indianapolis.
Before attending the Metis Data Science Bootcamp, Leon was a U.S. Air Force ROTC instructor teaching leadership and discipline. He also served as an Operations Research Analyst for the U.S. Air Force, where he used experimental design and hypothesis testing to guide strategy-level decisions. He was kind enough to share his time and answer questions from those interested in (or preparing for) the bootcamp, as well anyone else curious about his journey and story. Read the Q&A below, which includes specific questions about his time in the Air Force and his use of GI Bill funds to attend Metis.
Could you give a brief description of an average day of work as a data scientist at Viral Launch?
Luckily, I’m fortunate enough to have a pretty Python-heavy, abstract-discussion-based job at Viral Launch. I come into the office, check my email really quick, and then continue working on my current project. When I got the job, I had a few days of setting up my “stuff,” and as soon as he was able, the Chief Data Scientist gave me the rundown on a project he wanted me to work on, and I got working on it right away.
It already involves cleaning data and setting up a framework for modeling (so, in Python, I write code to clean data, read it into a machine learning modeling framework such that I can report results, etc.). I’m working in Python pretty much all day, and when I’m not, I’m talking with my supervisor about things he’d like me to do in the future or questions I have for the project at hand. We meet at 3pm to review each day.
How did you work with data in the Air Force?
I was an Operations Research Analyst and I worked with the Education and Training “department” of the branch to do studies and analyses on training data (along with some for the chaplain corps). For instance, if we train pilots using pipeline #1, will they do better than if we give them pipeline #2? Or, does this new tool for the nurse training school actually help students grasp the subject matter better? These simple questions can become pretty big data analysis problems, and it helped me to understand how data analysis plays into bigger organizations like the military. The work I did could be considered basic applied statistics.
In what ways does the Air Force use data science? And what tools are they using most?
I think that my perspective on data science in the military is a bit obstructed, or blurred. I was not really working in the ‘tip of the spear’ sections of the Air Force where a lot of the sexy machine learning was taking place. I can say that most of the data science being used (in these cool places) are being done by civilians or contractors.
The USAF hasn’t yet gotten to the point where there are active duty data scientists or ML engineers. I can also say that the new thing in acquisitions is definitely AI. I went to a talk about the three pivots of the USAF and the military (I can’t remember the first, but the second was jet propulsion around the Korean War, and third is AI). There are some cool Swarm Theory-based applications in the Air Force right now that I’m not super versed in, but I’m sure there’s a lot of data science involved.
As for tools, again, I can’t speak to the cool ML side, but in the data analytics side, pretty much everything we did was in MS Excel or SAS. I got suuuuper good at Excel. You could say I excelled in it. :)
What was the process like for using GI Bill for Metis?
The GI bill came to my attention right before Metis began. Already, I had planned on paying for it on my own (saving up, etc.). Once I found out about it, I got super excited. The process is very easy. When you separate/retire from the military, you’ll get access to a DoD portal where you can track your education benefits. You and (most likely) Leah (Director of Program Operations at Metis) will instantiate a sort of case file with the DoD, and first, you’ll get information about what your benefits are (if you didn’t already). This takes a few weeks. Then, you’ll set up your account information with both parties, and there will be a lump sum paid to you once it’s reviewed and approved. This takes about a month.
The amount that is covered under the GI bill for tuition goes straight to Metis (you don’t see that, and you’re billed for the difference, if there is one, based on your eligibility). Then, added in your lump sum are funds for room, board, and incidentals.
What was your best experience at the Metis Data Science Bootcamp?
I would say that the best experiences were the quintessential “Aha!” moments in class, where some esoteric concept became clear to me. Especially when I could apply it in a project, and was able to explain it to another student who asked about it.
I don’t think I could distill it down to one outstanding experience. The whole bootcamp was super rewarding and worth every moment.
Is there anything you wish you knew before going into the Metis program?
Going into Metis, I made sure to work as hard as I could on the pre-work they give to incoming students, and that helped me out a ton. But if there was anything I wish I’d known, it would probably be that I should really slice out more of my time to devote to the work given to me.
I am (was? am?) the type of person who (too often) puts more on himself than there should be, and I spread myself too thin. I wish I had gone into the experience with a bit more of a mindset of really committing as much time as possible to working on projects. (All things on the table, I had a small vacation planned during Metis…I should not have done that, haha!)
All in all, just make sure that your schedule is fully, FULLY open, and that you really plan to immerse yourself in the work. I wish I had done that a bit more than I did. (Even though I would still say I worked fairly hard on off-hours anyway.)
Were there resources that you found online that helped you complete the bootcamp pre-work?
I highly recommend Datacamp’s online courses, and frankly, Google. I would go through the pre-work with the mindset that you’ll have to teach it to someone else. If you come across a concept that you struggle with, Google it and try to find a good course on the subject matter.
I also found that writing and posting blogs from the standpoint of an expert (because you should be soon enough) on certain subjects in the pre-work helped. Finally, I highly recommend doing *all* of the work provided, including the bonus work and reading the books. You can skim through some of the books, but I recommend at least getting through half of the reading that’s provided. The more you get through the better. Eat it all up if you can.
What were some of the key aspects that you believe were pertinent to your success in getting hired closely following your graduation from Metis?
I would argue that the simple fact that I started applying a few months before Metis began was probably the largest contributing factor. I had my eye on getting an offer before Metis was through (farfetched, but a moon-reach goal in it’s right). I think with that mentality, and careful time management (this is key), I was able to find a steady footing in the job market at a good time.
It’s true, I have two degrees in Mathematics, but I think that the earlier you start to look, the better. Regardless of your background, someone out there really wants to hire all of you.
And one more thing, I recommend working on a few passion projects of your own. Conceive of some not-too-simple, not-too-hard data science project, and work to complete it (and maybe one or two others) before you’re inundated with Metis work. It’s nice to have multiple projects on your belt as you’re applying for jobs.
Do you have any advice related to the data science interview process?
All of those things you think that people consider when it comes to data science interviewing (what they wear, how they speak, how you respond to certain things, whether or not it’s clear you did your research, etc.) are in fact things that are considered. So, if someone (your career advisor, for instance) gives advice on how to go through the interview, I recommend heeding it. There really is A LOT that the HR people will think about.
Learn more about our 12-week data science bootcamp here, and see our events calendar for upcoming AMA sessions with grads now working in a variety of fields!