An interview with our Howest AI Engineer interns
Artificial Intelligence (AI) is an exciting and rapidly evolving field with numerous industrial applications. ML2Grow offers internships to students who want to deepen their AI knowledge and gain practical industry experience. We interviewed our two interns, Emiel Thys and Gerome Verzele, about their experiences with the AI Engineer course at Howest and what they learned at ML2Grow.
Q: Why did you choose the AI engineering course at Howest?
Emiel: “I am someone who wonders how things work. I am always curious and open to new information. This is why the course appealed to me. There were also a lot of opportunities, which reinforced my choice. Initially, I started with Game Development at DAE Howest. After two years, I set the bar a bit too high. The pace was very high, so I looked for an alternative, active and equally interesting course. ”
Gerome: “I have known for quite some time that I wanted to do something with computers, but which trajectory was not yet certain. Throughout the years at MCT (Multimedia and Creative Technologies), I got to know AI and started reading more about it, so I chose to specialize in it. I intend to have a bit of everything, but my main interest is AI.”
Q: How practical is the course, and are you satisfied with the quality?
Emiel: “In the course, the theory is often followed by a practical implementation or lab. This allows the knowledge to be applied in a supported environment. Things become clearer in this way, allowing for independent learning if necessary. As a result, I can be much more independent in dealing with combinations of known and unknown technologies. I am satisfied with the overall approach of the course. It provides an excellent foundation to work independently on possible projects in the future, provided with some starting support.”
Gerome: “The first half of the course is mainly focused on the design aspect, which was less interesting to me. But after choosing your “specialization,” the subjects became more interesting and better than the design subjects. Overall, it is very practically oriented, which is good, but sometimes you miss the theoretical part that gives you a deeper understanding of what is happening and why.”
Q: Which subjects do you feel are helpful during your internship, and why?
Emiel: “Machine Learning and Deep Learning are necessary to understand the general concept of AI. MLOps (Machine Learning Operations) also comes in handy to give clarity to setting up internal automation structures. Of course, new concepts are also discussed, but with the prior knowledge of these subjects, it allows for faster understanding and integration.”
Gerome: “Indeed. Most courses always come back a bit, so I couldn’t say which courses are not useful now and which are. One of the main technologies I’m using is Python, a programming language commonly used in AI and machine learning. I’m also working with libraries such as TensorFlow and PyTorch, which are used for building and training machine-learning models. Additionally, I’m working with Docker, a tool used for containerization, and Kubernetes, an open-source platform for managing containerized workloads and services. All these tools have been touched upon in our classes.”
Q: What is the internship project at ML2Grow?
Emiel: “The internship assignment revolves around segmenting point clouds. To create this, the focus was on implementing an existing model called RandLaNet (https://arxiv.org/abs/1911.11236). This model is certainly not the only one of its kind. The model related to the code made together with the paper is not built with recent libraries. This means the model still works on older and potentially less optimized code. Therefore, there was a need for an up-to-date model. The main focus was preparing a deployable container while reviewing possible data preprocessing.”
Gerome: “During the conversation with our mentor, it was said that there is a lot of freedom; if I find something interesting, I can certainly work on it, but occasionally they ask if you feel like working on something, and then you get a little assignment to work on. The nice thing about this internship is that you have freedom and don’t have to do the same thing constantly, so you can do all kinds of different things and learn a lot.”
Q: Do you feel like you are learning a lot during your internship period?
Emiel: “I think this experience will be very beneficial for our future career because it’s giving us a solid foundation in the practical application of machine learning in a production environment. By working on this project, I’ve gained experience with various tools and technologies widely used in the industry and learned how to apply them to real-world problems. Additionally, I’ve gained experience working in a team and collaborating with others, which is an important skill in any technical field.”
Gerome: “Yes, definitely. Due to the diversity of assignments you make, we come into contact with various technologies that are all interesting and make you learn. By working with these tools and technologies, I’m learning to build and deploy models, a critical skill for an AI engineer. Additionally, I’m getting exposure to real-world projects and challenges, which is helping me to develop problem-solving skills that will be useful in any technical field.”
Emiel: “In class, you are given perfectly executed examples ready for use. But this comes with a potential overflow of necessary steps never considered in the real world. The internship allows you to taste the real world/industry in a supported and controlled way. It shows that not everything is perfect or can be perfect.”