Road to Data Mastery
Part 1: Summer Intern
In this day and age, data drives businesses forward. With properly collected, sorted and analyzed data a good business idea can become a great business idea. We can find out the truth about what businesses are doing well and not so well, we can find out what will work and we can spread our story to others. Data science is super fascinating. I want in!
This is the first part in a series following my journey in becoming a fully fledged data scientist. My name is Emil Gunnberg Querat, 24 years old, I’m one year into a five year Computer Science Engineering Program at KTH in Stockholm. This first part will be about my first internship at a local TechFin start-up called Fairlo.
Leading up to the Summer Internship at Fairlo
I got in touch with CEO, Jimmy Hanna, in January, with the help of a friend at a local gym where I work part time and train at. When we first spoke on the phone he told me about Fairlo, the business idea, and we discussed what they needed help with.
Fairlo’s business idea is quite straightforward. Create financial services made for how people really live their lives. Fairlo’s first product is a realtime credit for people to even out their cash flow ups ‘n downs. Fully flexible and adjustable to each users lifestyle.
He went on rambling about data visualisation, compared self-written algorithms to using existing tools. He told me they wanted data from Google Analytics, Mixpanel Funnels, KlipFolio and a handful of services I had never heard of. A lot to take in, a lot to learn and a whole lot of excitement. I knew right away this guy is thinking big, and isn’t afraid of trying new things.
I heard they were planning on migrating to another database and thought they were expecting me to be part of that project. Holy sh*t. Okay, I better learn some database management too, before I show up in June, I told myself.
This was never the case. Thankfully… I was invited to meet them in the spring, have a chat, participate in a meeting with a data analyst they’d hired for some initial work. Right off the bat I felt a sense of openness at the organization. I asked questions, gave my input and was encouraged to continue! During the rest of spring, while finishing up the last courses for the semester at university, I started working on a case handed to me for training and I gained a much clearer idea of what I was expected to do during the summer. Fast forward to mid-June, I showed up, meet with the rest of the team and get to work. Finally!
What I learned about Data Science
Working with data science, although on an entry level still, I learned a thing or two about what the practice really means. I did learn plenty of SQL and how a database can be set up (yes, Fairlo actually let me do all the database queries so I had to learn SQL too), but data science is so much more than just the technical aspect of programming. Going in, I expected only to worry about how to query the database to give me the numbers that could make interesting graphs or neat charts. However, I was going to have to adopt a much deeper philosophy about how to deal with numbers and how to present them.
In order to become a successful data scientist, asking the right questions is more important than knowing how to implement data science techniques. What is absolutely central to the core business? What knowledge will propel the business forwards? And maybe most important, why do certain numbers and deviations happen?
As an example, displaying how the number of customers has developed since launch might hint something about how the business is doing. But, it will not necessarily give new insights about where to direct the organization going forwards.
A key point I’m taking with me from this summer internship is that whatever knowledge I want to dig out from the deep well of data should drive action. The analysis should grant understanding about the user, the marketing, the product or the procedure that inspire change or directs the development.
Relating back to the user count example, a more action-driven question to ask regarding clientele size is to explore how the profitability of a user is tied to rate of user inflow. Does high inflow carry less profitable users compared to low inflow, or vice versa, and what adjustments need to be made in either case?
This idea of data science spurring action was evident when I was given the opportunity to set up dashboards on television screens in the office. The team together discussed what is important to see daily and the central idea was that what is displayed on the TV screen has to be usable in our daily work. One example was to set up a display to show real time data on how our business is growing, for a quick and easy look-up when speaking with stakeholders and potential future colleagues. Another one was to see how our current customer base is doing, which might inspire a better customer service.
Looking ahead
After a summer of SQL querying, API requests and dashboard construction, I’m excited about the upcoming year of university. I will have a great opportunity to formalize my database management knowledge and learn new concepts of statistics alongside my continued part-time work at the up-and-coming TechFin star Fairlo.
