Tableau, Anomalies, and Friends: My Blibli Internship
It was the last day of my internship. I had a few things to do: send thank you emails, reset my work computer, and, of course, eat one last free chocolate-stuffed pastry from the company pantry. As I powered down my computer, I felt a hint of sadness. I was going to miss this place. I asked myself, “why will I miss this place?” I certainly was not going to miss the traffic on the way to work.
“Am I going to miss the projects?Processing millions of data points?” And then I realized: I was going to miss the people. Over the past few months, I worked with some of the brightest, kindest, and funnest (I don’t think that’s a word) people I’ve ever met. They were the ones who made this internship an amazing experience.
A little bit about the company and our team.
This summer, I worked at an Indonesian e-commerce company known as Blibli.com. I was able to work there thanks to the invitation of Johan Sulaiman, the head of Analytics at Blibli, and to the decision of Fritz Wijaya, the manager of the Analytics team.
Our tech team was small but versatile:
We had Gaby, Ivan, and Harris as the savvy Data Engineers, Ulil the full-stack-grandmaster-engineer, Joshua and Dennis the Data Engineering/platform/every-project-ever interns, Vincent and I as the fresh-from-America Data Science interns, Welly, the Data Wizard, and finally, Barry, the most overworked intern, who did QA/Testing for all of our projects. Our team was led by the fearless Johan Sulaiman and Fritz Wijaya, and supported by Willy our program manager and Ardy the project manager.
The lessons I learned.
At Blibli, I took part in many projects. My team created an automated ETL platform that extracted data from third-party APIs (like Facebook), cleaned 7,000,000 rows of data in Python and visualized it with Tableau, validated and created a clean customer master list in SQL, and built a time-series anomaly detection system to prevent fraud transactions. In addition to learning technical skills (thank you Stack Overflow), I learned many important life lessons that will help me to become a better Data Scientist.
If I was to explain all the things that I learned at Blibli, it would take about the same time as watching Die Hard 80,000 times. Instead, I’ve condensed the most important lessons into three points:
1. Collaboration is key.
“Collaboration is the key to success.” This is a phrase that we’ve heard many times from successful tech companies, software engineers, TED Talks, people on the street, etc. At Blibli, I was able to see this phrase transform from an abstract concept into concrete actions.
Blibli has an open-door policy, which means that any employee can talk to anyone. This policy creates a collaborative culture. For example, I saw my team work with the Digital Marketing team in Bangalore on a Customer Churn prediction project. I saw my mentor, Johan, proactively reach out to the marketing department to get their expert feedback on an ongoing Social Media Analytics project.
In my own projects, my success was due to the input and feedback of my co-workers. Every time I was stuck on a problem, my co-workers would gladly assist me (when they weren’t very busy, of course) and answer my questions. Without this culture of collaboration, my projects could not have been where they are now.
2. Results, Results, Results.
During my first week at work, my mentor gave me valuable advice: “Tim, every day, you need to produce results.” As a student, my job was to learn and understand theories, models, and technologies. However, at Blibli, I learned that in order for a company to thrive, I had to transform what I knew into real business solutions, and to communicate these solutions effectively. I had to be results-oriented and user-oriented.
This change of mindset was challenging for me. In school, I was used to taking the role of a researcher, focusing my efforts on perfecting my model, making sure that every step was perfect, and finding the best, most cutting-edge way to answer a question. However, at Blibli, I had to provide a solution in 1–2 weeks with limited resources! Furthermore, I found that my perfect, cutting-edge solution sometimes was not what the user/client wanted.
My experience at Blibli taught me to build solutions from the bottom-up. First, create a simple, relevant, and effective solution, then build complexity from there. Work with your users and managers. Make sure your goals are aligned. Strive to create a working solution, not a perfect one.
3. Learning Doesn’t Stop at Graduation.
Tableau, Google Cloud, Python, Docker. These are just a few of the technologies that I learned over the summer. Every single day, I learned something new.
As a student, I can have a week to learn a new topic. For example, let’s say that I’m learning about Decision Trees. That week, I can read books about the theory behind the model, understand the different algorithms that implement the model, and implement the most efficient algorithm to a homework problem I have.
In Blibli, I had to learn all of this in one day. There was no time! In order for the company to improve, its employees had to quickly move from learning new technology to implementing that technology. At Blibli, I became trained to learn efficiently.
In conclusion, I had an amazing time at Blibli! I was placed in a great team and learned so much from them. Blibli is a great company thanks to the hard-working and collaborative people in it. I hope that I will cross paths with Blibli sometime in the future, and I encourage those of you who are considering a career at Blibli to take the leap! In Blibli, you’ll find great technology, and most importantly, great people.
