A Banh Mi with Rafael



February 23, 2018

“We are WATTx” is our article series in which we draw back the curtain and give you a sneak peek into how it is to work here.

For this article, we went to the Cô Cô Deli to enjoy a delicious Banh Mi with Rafael Schultze-Kraft, our data science lead here at WATTx. We talked about his professional past in the field of neuroscience; his view on innovation, data science and people’s privacy; and lastly, what the future might hold for him, here at WATTx.

Hey Rafael! Tell us a bit about your professional past: what did you do before joining WATTx?

I have a background in Neuro- and Cognitive Science: I did my bachelor in Cognitive Science, and then I went on to study Computational Neuroscience for my Master at the Bernstein Center for Computational Neuroscience Berlin. I was intrigued by the big questions about the human nature no one had answers to, such as “How does the mind work?”

For those of us who don’t know: what’s the difference between Cognitive Science and Neuroscience?

Cognitive Science is the interdisciplinary approach to understanding the mind by bringing together various scientific fields such as Neuroscience, Philosophy of Mind, Psychology, Artificial Intelligence, and Computer Science.

Neuroscience is more focused towards the question: How does the brain work? And Computational Neuroscience, where I did my Master’s in, is even more specialized and technical: You create mathematical models of neural processes based on empirical data such as EEG signals in order to understand the underlying principles of the brain.

What brought you to Data Science?

The link between Computational Neuroscience and Data Science is that the methods that you learn by working with some particular data are essentially applicable to other kind of data as well. Data Analysis, Machine Learning and Data Modelling give you the means that are applicable to any domain. Whether you apply these tools to signals or data from the brain, to customer data, or to sensor data, the toolset will remain largely the same. In the end, you are dealing with just numbers. What changes is the domain of the data and the problem you are trying to solve - which is equally important to understand.

After obtaining my Master’s degree, I felt the urge to move into an environment that was more fast paced: Things in academia are usually slower and processes take a lot of time. I wanted to experience what I could do outside of academia.

You’ve been working for WATTx for two years now: How has your position evolved in that time?

It has changed a lot as WATTx has pivoted quite a few times. It took some time for us to find the sweet spot of where we wanted to go. And I guess we are still within this process, which I actually appreciate a lot, because most of the time, for me, change is associated with something positive.

We started off working within the main business of Viessmann, for instance working with sensor data from boilers and smart thermostats, which is very much related to their innovation plans. From there, we moved on and became more and more autonomous, turned into an independent venture builder, and are now also offering company building as a service (CBaaS).

Can you think of two examples that illustrate how different the work can be for a Data Scientist at WATTx?

For me, the Data Science space is essentially a spectrum that on one end is very analytics-heavy, from crunching numbers, analysing and visualizing data, extracting patterns and information, to building models and predicting something. The other end of the spectrum is more engineering-heavy. The beauty about WATTx is that our projects are very diverse and always cover different aspects of this spectrum!

A great example is the prototype we built for our venture Snuk, in which we analysed and modelled sensor data from office buildings, and developed automated and scalable data pipelines in the cloud to run these models.

Moreover, we also have research-heavy projects, like our venture Statice. With Statice, we are essentially building technology that can only be found in some recent papers from academia. It’s cutting-edge technology that is still in its infancy and where most of the work has been done in universities. For us, that means that we have to implement the information we get from research, and then build a data product around it.

As you can see, we have all these different approaches and problems: different in how you tackle them, and in the nature of the challenge and the problem itself.

And how does that compare with the previous jobs you had?

In many companies, work is centered around a particular product, so as a Data Scientist you are likely to work with similar data and problems over time.

At WATTx, this is very different, because we constantly work on multiple projects from different domains, with very different data and problems to solve. This gives you a lot of diversity in your everyday work, and provides you with the possibility to learn new things and technologies. Furthermore, we mostly build prototypes and MVPs, which is very different from working with systems in production. Finally, the work is extremely interdisciplinary - I work together with UX Researchers, Business Developers, and Engineers.

Data Science should always be attached to solving a particular problem.

What does it bring to data scientists to work in interdisciplinary teams?

In Data Science it is critical to combine data with an actual problem and an actual business case. This is one of the biggest mistakes that people make: doing Data Science just for the sake of doing it. Data Science should always be attached to solving a particular problem which is why I think the WATTx environment of having small interdisciplinary project teams really emphasizes this way of thinking and working together across departments.

Some people have a negative view on Data Science, seeing it as an instrument to manipulating people. Doesn’t this field come with a big risk of intrusion in people’s privacy?

I think we have a big responsibility to deal with people’s private data correctly and ethically. Technological advancement is enabling the creation of even more and more data and that data is being used to improve many areas, such as the health sector. Nevertheless, I think what needs to happen is that society and politics start catching up faster: There need to be laws and regulations that offer frameworks in which people’s privacy is protected. This is crucial.

Would you say that regulations today are lagging behind?

To some extent regulations are lagging behind, yes - in some places of the world more than in others. I think we can be lucky to live in Europe, where data usage is much more regulated than in the US for instance. With the GDPR and e-privacy regulations coming up, we see movement happening in this space.

There need to be laws and regulations that offer frameworks in which people’s privacy is protected.

There seems to be a trend towards Data Science in the industry right now. Do you have a possible explanation for that development?

I think that the field is popular because of the enormous amount of data that we have at hand today. People start to realise that they can make use of data for their businesses and for driving decisions. In addition, this amount of available data together with advancements in computer hardware is enabling state-of-the-art technology such as deep learning, and people are noticing how much is possible using these technologies.

How do you see the future of DS?

I think Data Science is evolving: A few years ago it was hard for many people to easily build machine learning models because of a lack of easy-to-use tools and much more theoretical background was needed to develop those. Nowadays, information, tools, and open-source code are available all over the place and with those at hand, Data Science is somewhat less technical and becoming more multifaceted; including asking the right questions, understanding the business value, and knowing to appropriately apply the right tools for the data at hand.

How much utility do we get out of data currently?

The potential is huge. But the question is tied to how many people actually make use of it. And I think the answer is: still way too few, especially within larger companies in Germany. There is so much that can be done. Many companies today collect and store all data available, but then they don’t do much with it. Why? Because they don’t know how to do it. There is a lack of know-how in the industry, and a lack of data scientists. In start-ups that live in data- and innovation-driven environments, like the one we have here at WATTx, things of course are different, because the potential of data is being recognised and leveraged.

What would you say to people who apply for the data scientist position at WATTx?

One of the great things that we offer is the diversity and flexibility in the work. That is why we like to look for candidates that are generalists: Curious and open people with a willingness to learn, touch many different technologies, tackle many different data problems, and are curious about working on the full spectrum from Data Science: from high-level analytics to more engineering-heavy tasks. It’s really a great environment for self-development and dealing with a broad range of challenges in the deep tech space.

One of the great things that we offer is this diversity and flexibility in the work.

What is the most difficult part of your job at WATTx?

Maybe wanting to do everything at the same time: There are so many interesting conversations, so many ideas and so many things happening at the same time, and I want to work a hundred percent of the time on each of them - which of course is not possible.

What is the most rewarding part of your job?

Many things! Seeing companies being spinned-off that are based on our hard work, delivering successful prototypes and MVPs, seeing the huge interest from the industry approaching WATTx, and noticing how much we are learning along the way. But above all, working together with an amazing, motivating, and highly-skilled team!

Do you have a favorite moment at WATTx?

I have to say that I really like the small moments a lot. For instance friday evenings, when we get together and play board games with a beer. Or our hackathons. Or enthusiastic conversations around new ideas. These moments I really love. These occasions show who we really are.