Mon Jan 16 2023

Xpert Insight: My experience at Xomnia’s Data & ML Engineering Program

By Anna Dekker

NB: This blog discusses a program that Xomnia no longer offers.

My name is Anna Dekker, an alumna of Xomnia’s Data and Machine Learning Engineering Program. After completing the year-long program, I joined Xomnia’s team as a data engineer. Below, I share my experience at the development program.

A little bit about myself...

After my graduation with a masters degree in econometrics from Erasmus University, I realized that I wanted to continue developing myself on a technical level, but also contribute to addressing practical problems for companies and institutions. It felt as if this came perfectly together in the realm of data science, which led me to is why I started looking for a data science position.

While working as a data scientist, I enjoyed wrangling data and building models, but I felt that I missed a part of the puzzle as I had no idea how to productionize my models. I wondered how to automatically feed new data into my models, what infrastructure was needed to achieve this, and how to scale for more data. Consequently, I found myself reading blogs on sites such as Medium and Towards Data Science about all these data engineering tools and concepts that I had no experience with, but really liked to know better.

At the time, I coincidentally came across a blog on LinkedIn by a former trainee of the Machine Learning Program at Xomnia (now known as the Data and Machine Learning Program). It was as if I was reading my own words, because there were many similarities between his story and mine (he studied econometrics as well and also became interested in data engineering after starting in data science).

Soon I decided to apply for the MLE program, and one year later, here we are!

We're hiring machine learning engineers!

My experience at the Data & ML Engineering Program

In the program, my week was divided between four days of working at a client and one day dedicated to training on topics in data engineering and machine learning at Xomnia.

The program has given me a good foundation on what to look out for when applying certain techniques or concepts, and which intricacies are difficult. There have been numerous situations at the client that involved concepts that in some manner had been part of the program. This often gave me a head-start, as I knew because of the program to search in which direction.

What I especially liked about the program is that all training courses are taught by experienced Xomnians, who know the ins and outs of the matters they’re instructing about. This is particularly helpful since they could share first-hand experience about how they have mastered a particular skill, and the pain points that they have encountered.

Looking back to the start of the program, it is fun to see that many things I wondered about back then have been answered, but many new questions have popped up, as there are still many subjects I would like to master further!

While working at the client

During the program, I worked at Bindinc, one of the largest media publishers in the Netherlands. Besides their activities on publishing TV-guides, Bindinc is one of the largest players in video-on-demand data. I contributed to developing and maintaining their ETL pipelines. I also helped re-designing their cloud infrastructure to allow for a more scalable and robust cloud architecture, leveraging the benefits of the cloud.

Being able to apply what I learned in the training workshops at Xomnia during my work at Bindinc was very valuable. This is because getting the code to work through all bugs really supported my learning experience.

Currently, I am part of a project at the Municipality of Amsterdam in a team that mainly focuses on infrastructure deployment. Together with my colleagues, I build pipelines to deploy robust and secure cloud infrastructure that can be used to host APIs. Because it is quite an intricate infrastructure, we have to ask ourselves some interesting questions, like how to program the pipeline in a way that keeps everything neat and easy to interpret, and what set-up is best for our CI/CD? Moreover, it is fun to dive into all the cloud functionalities, from networking to compute, to explore and apply the settings that fit our use case best.

Why work for Xomnia?

The program gave me a full-tour around all major subjects in data and machine learning engineering, and I feel that I have become a much more well-rounded engineer.

It feels very empowering to become more vocal in engineering, because it is so exciting to see that you can contribute to many meaningful goals that can be achieved using a form of coding or automation. For example, it was very inspiring to participate in a coding challenge organized by Fruit Punch AI to support wildlife conservation in South Africa. Together with fellow team members, I developed an RGB model to detect rhino poachers.

Besides the theoretical side, the interaction with a large pool of engineers has been inspiring. Xomnia regularly organizes expert talks, in which colleagues share about a topic they have strong affinity with. I also frequently discussed with a senior colleague on how to implement a specific feature or get through a certain bug. I sometimes also received advice on ‘unknown unknowns’ to improve a code that I was not even aware about. This definitely steepened my learning curve, and I have really enjoyed being able to share insights and to learn with each other.

Group photo upon completing the program with the graduating batch