Engineering and Product
Muriel Hol, Lead Data Scientist at Halter
A move from the Netherlands to come and cuddle our cows, Muriel Hol, Lead Data Scientist at Halter has had a fascinating journey into working in tech.
A dream of one day opening a restaurant was thwarted when falling in love with artificial intelligence to which she now applies to understanding the behaviour patterns of cows. Great to have you in New Zealand Muriel and thanks for sharing your story.
“Imagine watching 200 cows at the same time for 24 hours. You would get dizzy! ”
Firstly, how would you explain to a five year old what it is you do?
We collect a lot of information from the cows, how they move, when they eat, where they are. And this is great, because we, data scientists, believe that the more information you can process, the more you can learn from it, such as when the cow is sick.
Unfortunately, we humans are not good at processing a lot of information. Imagine watching 200 cows at the same time for 24 hours. You would get dizzy!
Luckily, computers can easily process a lot of information to help us out, and data scientists know how to explain to the computer how to do that.
And for the adults, what does that translate to in regards to your day-to-day?
Firstly, a data scientist has to get extremely familiar with the data; e.g. how, when and why does the data change over time. This information is crucial to be able to get insights from it using statistics and expert knowledge. This means doing literature studies, being in the field (in our case, being on farm) and making different insightful visualisations of the data.
Secondly, it is building models. Either by writing up the rules that use those statistics to make decisions, or to make machine learning models that learn how to make the decisions from the data. Lastly, it is important to continuously evaluate the models, both the performance on the data and the value they bring to the customer.
What are some of the common misconceptions about working in Data Science?
A quote from a recent Google-paper sums it up: “Everyone wants to do the model work, not the data work”. Without high-quality data, data science can not succeed. Data science is not only about maths and machine learning, it is just as much about knowing how to get the data that is needed. In my experience, this is the hard, but most important part. Something that is often overlooked.
“..opening a restaurant has been my dream. Until I started my first course in artificial intelligence, I was completely sold.
Was working in tech something you dreamed about doing as a kid, if not what was?
All-tough I always enjoyed science, I only discovered I wanted to work in tech when I started studying AI. As a kid I mainly loved being on my feet, being busy.
My mom had a second-hand toy store, where I used to help as a kid. In the weekend I helped out at my horse riding school. I loved feeling useful, and being out and about. I think that is why I loved being a waitress for a long time, and opening a restaurant has been my dream. Until I started my first course in artificial intelligence, I was completely sold.
Tell us a little bit more about your career journey and ultimately about how you ended up working at Halter?
During my masters I was a teaching assistant for the bachelor artificial intelligence and I also worked part-time at a fin-tech company, where I learned about the tech behind the payment industry. I became passionate about time-series prediction and did an internship at a crypto-currency trading company to research this.
Here I learned that it is extremely hard to translate what I had learned in uni to real-life applications, which was a very valuable lesson. After my internship it was time to write my master thesis. A week or so after graduating I emailed Halter.
I knew I wanted to do something with sustainability and AI. After some research, I found out about Halter. The mission and the idea of building models while working with animals, sounded like a dream. Definitely worth moving to the other side of the world!
What is the best piece of career advice you have ever received?
“Own your career”. Explained well by this quote by Toni Morrison; “Whatever the work is, do it well—not for the boss but for yourself. You make the job; it doesn’t make you.”
“Also, being able to cuddle a cow once in a while is a dream coming true.”
What do you love about working at Halter?
That mixture of creative freedom, ownership, and the speed at which it needs to go. It’s just a great train to be on. I haven't had a day not wanting to get on it. Especially with so many smart people in it, you learn a lot. Not only in engineering, but I have learned just as much from how people deal with life, from how kind they are and how hard they work.
Also, being able to cuddle a cow once in a while is a dream coming true.
Lastly, Halter continues to grow and evolve. What kind of candidates do you think Halter is looking for in terms of experience, attitude and character?
As things are being built fast, I think it is important that you can embrace the need-for-speed while also working data-driven making decisions objectively. Because of the complexity of the problems that need to be solved, you need to be someone that does not get frustrated but that is persistent and optimistic, while also being able to let go of projects when needed to try a different path.
Lastly, in my experience, being in the field, taking interest about farmers and the complex and serious problems they need to solve on a daily basis, is invaluable to understand your job. You need to care about the mission, care about the farmers, care about the animals.