Interning at Blibli.com was a really memorable and rewarding experience for me. Over the past month I experienced a lot, met many new people and got to work on a challenging yet fun project during my internship — one that could significantly impact the Blibli.com’s user experience in the future.
A few things jumped out to me on my first day at Blibli.com — the company (and the Grha Niaga Thamrin office) was much larger than what I thought it’d be. There were so many people in all the different parts of the office, all with open-plan style tables. Everyone had a pretty casual culture, with casual pronouns often thrown in in conversation — but everyone also always gets serious when and where it counts.

Some time after I arrived, I was given a tour around the office and was given an introduction on how Blibli.com runs and the team I would be working with. I then finally got to meet Andi, the R&D lead at Blibli.com in person. From there I then went on to meet the rest of my colleagues later on: Hendri, lead of the Machine Learning team in the R&D division, Uriando, and Wiega (a fellow intern).
It was clear that they all shared the same passion from the discussions we had and how we worked together and got the project moving well on its way. I was fascinated and excited to be able to both learn new concepts and skills and take into practice the skills I learned at uni.
The project was not easy — we often ran into snags, with our model behaving unexpectedly due to things like a numerical underflow problem we ran into. I also had to learn some additional background knowledge in mathematics, probability and statistics — new concepts like mixture models, likelihood, Bayesian models and many others, to get to a point where I could fully understand Hidden Markov Models (HMMs) and implement it well.

As the famous saying says, “time flies when you’re having fun”. Between all the rounds of ping-pong at lunch break, the new people I got to meet at Blibli.com, and all the work me and my team did, my time interning at Blibli.com soon came to a close. We hosted a sharing session on HMMs, where Wiega, Uriando, I and Hendri presented the progress in our work with Hidden Markov Models and fielded some questions from people interested in using the model. In that session, we went and briefly explained Hidden Markov Models — what they are, how they work, briefly going through a few of the papers, and even had a demo of a HMM-based recommendation system in action for the audience. We then had a final lunch the next day at D’Cost Thamrin City. I really can’t say that I didn’t get a bit sentimental when I signed off for the last time that day.
For me, Blibli.com was a really great place to work — and if you have not considered Blibli.com as a place you would love working at, give it some serious thought! I believe and truly wish Blibli.com success in achieving their goal of becoming the №1 e-commerce in Indonesia!

