During his studies, Marcel developed a keen interest for spectroscopy and chemometrics. He was especially curious about data visualization and descriptive data analysis. So, after completing his master’s degree in chemical biology and starting his PhD thesis, Marcel joined Senorics in 2018 as data scientist.
He has taken the time to give some insight into his work at the company.
What led you to join Senorics?
“During my studies, I also worked at the university. Eventually my contract was running out and I was looking for a new challenge. Working at a startup seemed like a good idea, as it offered the right balance between academia and industry. Senorics had just been founded and I was intrigued by the prospect of being part of this endeavor right from the beginning."
Marcel Dahms, Data Scientist
What are your main tasks and responsibilities as data scientist?
“I’m the team leader of the data science team, so one of my tasks is to do the strategic planning, assign tasks to all the team members, keep an eye on ongoing projects, and align with other team leaders about current topics.
One of my daily tasks is to conduct data analysis and consult with the sales team and customers to align on use case and application requirements. I build chemometric models for those customers, so that they can benefit from Material Sensing. This includes planning measurements, organizing samples, and testing the model once it is built.”
What is the general process when developing a new application? Do you follow a certain set of steps?
“I do follow a certain set of steps but describing them in detail would probably go beyond the scope of this interview. To summarize, it’s important to clarify customer requirements before starting to build the application. It’s also important to test the feasibility of the customer requirements by conducting a proof of concept. Once this is done, I conduct the first measurements and source reference values. With this data, the chemometric model can be trained. Following this step, the model is tested with new samples and optimized if necessary. Finally, the model is tested by the customer in the field and if no further optimization is necessary, the application is released.”
What do you find most challenging in your job?
“Chemometrics is very complex and so it isn’t always easy to determine what kind of chemometric model is most suitable for the customer. What may seem like insignificant details, can have a major influence on the development of the model. So sometimes it takes a little while to determine all the requirements before an application can be built.”
What do you find most rewarding in your job?
“Seeing our devices being used is very rewarding. It’s nice to enable customers to benefit from our technology. It’s also very satisfying to work on a model for several weeks and then the moment arrives when it is working as intended.”
We would like to thank Marcel for taking the time to answer our questions.
Within the next few months, a new episode of the SenoTalk will be released, in which Marcel dives deeper into how applications are built and shares insights into the projects he has been working on. Stay tuned and in the meantime take a look at our previous SenoTalk episodes on YouTube.
