The latest news from The Hyve on Open Source solutions for bioinformatics

Recent Posts

Hyver Stories: Software Architect Gijs Kant

January 13, 2020 | By Gijs Kant

Gijs portrait colorGijs Kant (38) joined The Hyve in February 2015.

Since then, he has been involved in a variety of projects using tranSMART. More recently, he contributed to the Fairspace project for the Curie Research Institute in France. Next to his job as a Software Architect, Gijs is working on his PhD thesis about verification of the correctness of systems based on formal logics and games using high-performance algorithms. In his spare time, he likes to play the organ or cook for his family.


 

 


Can you tell a bit about your background?
I started my studies in Business and Information Technology at the University of Twente in Enschede. After a while though, I found the course not in-depth enough with regards to the technical aspects. So after two years I switched to Computer Science.
After finishing my studies, I decided to do a PhD. For 4,5 years I was immersed in research, and I liked it very much. By the end of my PhD contract, I had finished the research part but not the writing-up part. It’s something I’m still working on. With a full-time job and three kids at home, the writing is not progressing as quickly as I’d like but I am planning on finishing it.

 

How did you get to know The Hyve?
By the end of my PhD, I started looking out for an area where I could apply my knowledge. A brother-in-law suggested I’d look into The Hyve. Indeed, the company appealed to me as it has the medical sciences as its area of application on the one hand. And it allowed me to stay in touch with research on the other hand, by developing software applications for scientists. I also liked the open-source approach. The field is still very much in development and there is enough room for improvement. So I felt it would be nice to contribute to that.

computer-work Gijs

Did your PhD research prove useful for your work at The Hyve?
Certainly. At The Hyve, we regularly encounter performance issues. My research taught me how to systematically analyse a system and identify factors that contribute to a reliable and fast system that doesn’t require too much memory. That’s useful knowledge when you want to make sure that a system is not only working in a test environment, but is still fast enough when scaling up to, say, 1000 users.

Which team and products are you involved with at The Hyve?
I used to be part of the Data Warehousing team. That team is currently being merged with the Fairspace team. It means that suddenly we’re responsible for a much wider range of products. It also makes it a bit difficult to say what my work entails, as a number of projects I’ve worked on are being wrapped up and we’re only just starting new ones.
A major customer I’m involved with at the moment is Institut Curie in Paris. We’re developing solutions for its data and file collections storage. An important aspect of the project is the addition of metadata. With metadata you can capture a broad range of characteristics: descriptions, data type, provenance, disease code, patient identifier, et cetera. You want to lay down certain rules for which metadata is captured and how, otherwise you can’t retrieve the relevant data when you search for them.
At the same time, you want the metadata model to be flexible, so you can add relevant fields at a later date. We capture metadata using a Resource Description Framework (RDF) data model. Working with a standard guarantees easy coupling of data and metadata, the option to link to other systems, and allows you to easily retrieve data.

What do you do at The Hyve?
As a software architect, I’m usually working with a small team of developers. Together we try to come up with solutions for the IT problems our clients encounter. I like the conversations I have with clients about the data they have, the number of users, the solutions they need, and things like that. I also like the design aspects of the job: figuring out with the developers which open-source tools to use and how to adapt these for the client’s IT infrastructure.
I think it’s important to build systems in such a way that you can recognize separate structures and reuse the code. For example, the different functions of an application should be contained in separate components that are linked together in a logical way. This enables you to reuse the code when you build a similar system that only differs in certain aspects.
Of course, I also want to make sure that the systems we deliver perform well. That means testing them thoroughly before implementation, so you are confident that everything works well. Since we are working with open-source systems, another factor we need to keep in mind is that our adaptations should not only work for a particular client but also for other organisations.

Gijs talking at the Tübingen TranSMART-i2b2 meeting October 2019

Gijs talking at the Tübingen tranSMART-i2b2 meeting October 2019

 

What do you like about working at The Hyve?
The colleagues are nice, so it’s always nice to work together on a project. I also like that there is room for experimentation. At the start of a project, it’s not always set in stone which technology we’ll use. Of course, it should be clear if a certain goal is achievable. But if someone wants to try out a novel technology, there is often the opportunity to give it a shot.
At the same time, we don’t dismiss proven and trusted techniques because at the end of the day, we want our products to be stable and reliable. But I like the atmosphere of curiosity and interest for new things that you find across the entire organisation. Something that is also very much ingrained in the company is support for the open-source model and the drive to build good tools for biomedical scientists.

Can you mention an exciting development in your field?
An interesting development is that more and more services are offered in the Cloud. Big players, such as Amazon, Google and Microsoft, say: leave it with us and we’ll take care of it. That applies to databases, data storage and hosting all kinds of systems. For organisations, may sound like an attractive and cost-effective scheme. But a big disadvantage for developers is that you lose quite a bit of control. You don’t know exactly what goes on under the bonnet and it makes you dependent on these providers. Besides, you need to host all data and systems with the same supplier.
A more positive development is the implementation of standards with regards to data exchange. When everyone uses the same standards, you can link systems and exchange data and it doesn’t really matter which systems the various organisations use. This is a huge advantage in the medical field, where hospital, GP and pharmacy systems need to work together.

Gijs at organWhat do you like to do when you’re off duty?
As I mentioned before, I have three kids. They demand a lot of my time in the evenings and during the weekend. Also, I take the kids to school before I come to the office.
My biggest hobby is playing the organ. I like to rehearse classical music scores, as well as listening to classical music. Once every fortnight I play the organ in church.
I also like to cook; you’ll often find me in the kitchen during the weekend.