Asking Alumni #2
Fri, 24 Jul 2020
written by Jorrit Geels
As a student in one of the most widely applicable fields, it is easy to lose yourself in the amount of options you are given. But why figure it out all by yourself? There are always others sharing the same struggles, both now and in the past.
I decided to reach out for advice to those who have been able to deal with my struggle in the past, our alumni. These kind people took their time to review their career for me: the choices they made, the problems they had to deal with, and what in the end made it all worth it.
The format is very simple: they all answer the same 10 questions in their own way. No guidelines, no restrictions: their story.
If you are/know an alumnus/alumna (preferably 5+ years since graduating) that would like to participate in these blogs, please email me at cogblog@svcognac.nl.
On https://svcognac.nl/blog/132 you can find the first four interviews. Today, I present you the next four I received. A huge thanks to Jeroen, Wouter, Bart, and Maya for being a part of this blog series.
I’m curious to hear what you, the reader, take away from these interviews. Please let me know through cogblog@svcognac.nl. Some will be featured on the next post (around October), anonymously if you prefer.
Have a nice summer.
- Jorrit Geels
Contents:
- Jeroen: Autonomous robotic systems developer at the Royal Dutch Army
- Wouter Bulten: PhD candidate at Radboudumc
- Bart van Delft: Software engineer for YouTube
- Maya Sappelli: Applied Data Science researcher at the HAN
Jeroen
Jeroen started his studies on artificial intelligence 15 years ago. After having worked for ING in the Netherlands and Australia, starting his own business, he now develops autonomous robotic systems for the Royal Dutch Army.
His biggest accomplishment though: “I asked the love of my life to marry me, and she said “Siiii!””
University and CognAC
1. Were you involved within CognAC? If so, how? What did you learn from that experience?
I was in the Activiteiten-Commissie (AC) for two years, organised our Batavierenrace participation with team ‘InTenShardmetCognAC’ three times, did the KasCo on a blue Monday (thanks Rik!), and organised several other activities here and there.
Of course I gained a lot of xp in organising and planning, but probably the most valuable thing I got out of it was getting to know my fellow students a whole lot better. It made my student-life a lot more active and fun.
2. What is your most ‘typical CognAC’ memory?
The TK (Terminal Kamer) was just a special place, where hard work was combined with lots of socializing and games of multiplayer frozen bubbles on a single keyboard.
3. What is/are the most valuable thing(s) you learned in university, outside from course material?
- Working systematically: Our professors always gave us clearly-scoped assignments, and taught us systematic ways to approach the problem. Real-life is not like that, it is volatile, uncertain, complex and ambiguous and there is no one to tell you the right approach.
- Reflection. Or maybe more generic: The ability to learn new ideas and concepts. My grandfather used to say “je moet leren leren” (“you have to learn how to learn”), and ofcourse -as wise as he was-, he was right.
About your career
4. What is it exactly that you do now on a daily basis?
I work for the robot unit of the Royal Dutch Army where I both manage and actively participate in the development of robotic units that will help soldiers in the field.
Half of my time I work with TNO colleagues on projects that could help the army in the future. Among others we look at how an autonomous system could reason in the field with hierarchical task networks.
The other half of my time I have the flexibility to work on projects for the unit directly. Among others we are working on a project with which we can control robotic units through gestures.
All in all it’s a pretty cool job :)
5. What skills are most valuable in order to perform well in your workfield? What would you advise students to acquire these skills?
Honestly? Invest your time in the skills you find interesting, together with the skills your professors are trying to teach you (turns out, they will come in handy).
Together they will give you a combination of skills that are useful, and that you are passionate about. Almost every other skill you need in the workfield you can learn on the job (imho).
6. Are there any things you wish you’d have done in your time as a student?
Tough question. I would have liked to be a student-assistant, joined the study trip to Sweden, followed several extra classes, joined Phocas and quite a bit more.
My student time was an awesome time, in which I learned a lot in the classes, but definitely also outside. All the extra activities taught me a lot about myself, developed basic skills like planning, organising, people skills, etc. Looking back, I wish I did even a bit more.
Dealing with problems
7. What is the hardest thing about your current job? How are you dealing with that?
The balance between management and my own contributions will always remain tough. Managing your own tasks, emails, projects, and colleagues is always ad-hoc work, with lots of context-switching. To contribute on a project you need the opposite (especially as a developer): focused time in which you can deep-dive in to your code or problem.
I use my own self-developed task-management tool, but also resort to simple tricks like blocking my calendar and turning off all notifications for a day.
8. How did you come to your current job? Did you take steps in between? Could you consider any of these steps as ‘failures’, and if so, in what sense?
I took some steps to get where I am right now, but I learned many valuable things along the way.
After finishing my Master, I joined ING for a traineeship. I saw there the huge gap that exists between the material taught in university and the business world. It felt like I knew nothing, and had to learn everything from scratch. It was an incredible steep curve, and I gained a lot of experience in a short period.
I quit ING to pursue an idea of myself. I had always struggled with the many tasks I had in my life, and I had developed a tool that helped me manage that. It worked well for me, so I figured I could put this online and help other people. I thought the idea was simple enough to do everything by myself, but I had however (greatly) underestimated the disciplines required to put an actual product on the market (design, UX, software building, testing and releasing, marketing and business to name a few). In the end it cost me too much time to finish it and I stopped working on it full-time. The experience taught me a lot though, and the product still helps myself, my girlfriend and a handful of others on a daily base. This is something I'm proud of and still motivates me every day.
During that project I heard about an opening at an innovation role at the army. I had been a reservist there since I started studying and now there was a chance to make it my full-time job. At the job interview the colonel that was interviewing me asked me what was in ‘my toolbox’. I name some of the languages and frameworks I had worked with, but he quickly interrupted me: “Give me an example”. That is when I told him about my raspberry-pi weekend project with which I tracked when my roommates were in the home. The colonel loved that kind of my-work-is-my-hobby approach and I was hired on the spot.
From the innovation unit, I moved to the robotics unit where I am working today. Without my side experiences (as a reservist, and my own technical hobby projects) I definitely would not have ended up at this pretty awesome job.
Advice, to close off
9. What is the best advice you ever got? Why did it turn out to be so useful?
Stand up for your own career. Whether you are ambitious or not, it is important that you work on things you find exciting. You might have an idea what you find exciting, or maybe you are still trying to figure that out, just don’t settle for anything less.
If I would have followed every suggestion people would have told me (‘you are good at X, you should become our next X-person’), I would not have ended up in a job that I love so much.
It is like Steve Jobs said: "Have the courage to follow your heart and intuition. They somehow already know what you truly want to become. Everything else is secondary."
10. Do you have any good advice for ambitious students that would like to go into your workfield? And what kind of advice should they ignore?
Pick up a side project that you are passionate about. I learned a lot from these hobby projects, and somehow the experience always comes back useful in my working life.
I always experienced that learning is way easier when you do it without deadlines, or people telling you what to do, just do it for yourself.
Wouter Bulten
Even though it’s only been five years since Wouter Bulten graduated from Radboud University, he has already left quite a fingerprint in our field. He first worked as a programmer and data scientist before starting as a PhD candidate at Radboudumc in the pathology department. Keep on reading to find out more about the career of the 20th Chair of CognAC.
University and CognAC
1. Were you involved within CognAC? If so, how? What did you learn from that experience?
I've been quite active within CognAC in various roles. It all started with joining the 1C (First years committee) and the SymCo (Symposium Committee) in my first year. After that, it was difficult to stop!
The highlight was, of course, the privilege of being chairman of the 20th board of CognAC. In terms of experience, participating in committees and/or being part of the board is an excellent way of getting more experienced with having (efficient) meetings. Also, it's a nice way of getting more familiar with organizing events and project planning. Finally and most importantly, it's also fun of course!
2. What is your most ‘typical CognAC’ memory?
There are many to mention. Some highlights: the annual Waal-beach-campfire during the orientation week; they day-before-stress of organizing the symposium; the yearly constitution drink. In general, CognAC always felt like a big friendly family. RJ (treasurer of the 20th board) and I still try to go to the constitution drink every year!
3. What is/are the most valuable thing(s) you learned in university, outside from course material?
I've done quite some things for the university (e.g., faculty student council and faculty assessor). In these roles, you can learn how the university and a large organization works from the inside. If you continue in research, this knowledge can help with understanding things like the money flow, how the academic system is set up, how decisions are made, etc. If you are ever in a position to try one of these things: I can recommend it!
About your career
4. What is it exactly that you do now on a daily basis?
I'm working as a PhD candidate at Radboudumc in the pathology department. My research focusses on building AI systems for the diagnosis of prostate cancer. If you want to know more, I wrote a small blog post on our main study: https://www.wouterbulten.nl/blog/tech/automated-gleason-grading-deep-learning/
What does this mean on a daily basis? Well, most AI systems need a lot of data, so a large part of my work focusses on getting the data and transforming it in such a way that we can use it to train deep learning models. My biggest project so far, in which we showed that AI can outperform doctors in diagnosing prostate cancer, was probably 75% focussed on data. At the start, this data was not yet digital but was just fixated tissue stored in our archives. It took quite some time to get this data, digitize it, and train the deep learning system. The same holds for the Kaggle competition we organized (https://www.kaggle.com/c/prostate-cancer-grade-assessment), a lot of time is spent on getting the data. Apart from this, you don't perform science alone, so a large part of my work also consists of having meetings, discussions, and presenting my work to others.
5. What skills are most valuable in order to perform well in your workfield? What would you advise students to acquire these skills?
If you want to continue in research (but also other fields), good presenting, writing, and project planning skills are key. If you go to (international) conferences, it can be quite hard to give a captivating presentation and keep your audience interested. Especially if you're the 10th presentation they have seen that day! Regarding writing, it can save a lot of time if you're experienced with creating write-ups of your work. You have to write papers and a thesis after all if you want to get your PhD.
Project planning is useful in general, especially given the limited time you have in a PhD (yes, those 4 years are over before you know it). Of course, these skills are not limited to an academic career.
The best way to train these skills is to do them. For example, try to present as often as you can, especially if you don't like it or get nervous. One tip I received from a presenting course really captures this point: "Remember the 5 P's of getting good at presenting: practice, practice, practice, practice, practice."
6. Are there any things you wish you’d have done in your time as a student?
Difficult question! Maybe, in hindsight, I would also have followed Statistical Machine learning. For the rest, I think I'm quite satisfied with how everything went!
Dealing with problems
7. What is the hardest thing about your current job? How are you dealing with that?
A PhD is not a typical 9 to 5 job, and you have a lot of freedom and responsibilities at the same time. In the end, I am mainly responsible for getting my PhD done in four years. This means that you have a lot of ownership, which is nice, but if you encounter setbacks, your also the person that has to solve them. This can result in a feeling of being always busy. Dealing with this is a learning process. My way of approaching this so far is making sure you have a good team and try to often reflect on the process (and not only on results).
8. How did you come to your current job? Did you take steps in between? Could you consider any of these steps as ‘failures’, and if so, in what sense?
I always wanted to do a PhD after my masters. However, I first worked as a programmer and data scientist before I started my PhD. Someone in my network shared the vacancy for my current PhD project, so it was kind of by accident that I joined this project. Working as a programmer before doing my PhD was a nice experience and also helps me during my PhD. There is quite a gap between programming for fun or a course, and programming for a production environment.
Advice, to close off
9. What is the best advice you ever got? Why did it turn out to be so useful?
Not sure if it's the best advice I have ever received, but something that I definitely remember: Pim Haselager once told me never to end my presentation with just a question mark or a "Questions?" slide. That's the slide that will be present during the discussion and serves no purpose if it's just a question mark. Since then, I always try to have something meaningful there. Maybe a call to action, or a take-home message. I have been using that ever since.
10. Do you have any good advice for ambitious students that would like to go into your workfield? And what kind of advice should they ignore?
The barrier between research and industry is smaller than you might think. Starting a PhD project after your masters does not mean that you will work in research for the rest of your life. You can also start later, and especially if your PhD has an engineering component, this can even be good!
Moreover, most PhD students will work outside of academic research after finishing their PhD. Research is like a pyramid, and the higher you get, the fewer positions there are. From a personal perspective, this is something to think about before you start and during your PhD. What is your goal? Do you want to continue in research, or not? Does the topic you apply for match your goal? Are you able to learn skills during your PhD that can also be beneficial outside of academia?
Editor’s note: as last time, I tried to squeeze in some of the ‘did you know…’-like questions into the small introduction, but failed. Here are Wouter’s answers to the additional questions!
11. Biggest achievement: We showed that an AI system can outperform pathologists in diagnosing prostate cancer. This research resulted in various news interviews, which was quite cool!
12. Unusual hobby, habit, or character trait: I am a huge fan of home automation. If you ever need a new hobby, check out Home Assistant (https://www.home-assistant.io/).
Bart van Delft
After finishing his Bachelor in AI in 2009, Bart van Delft did a Master in Computing Science, focusing on Computer Security. During his PhD in Sweden he did a summer internship at Google, where he currently functions as a software engineer for YouTube. The fun facts that Bart answered for me can be summarized in “improvise, adapt, overcome”: he respectively likes adding musical improvisation to improvisation theatre, he started life in a new country twice, and he finally got his driving license last year after the first few tries during his Bachelor. At the age of 32, Bart has already gathered a lot of wisdom, which he will share with you in this interview.
University and CognAC
1. Were you involved within CognAC? If so, how? What did you learn from that experience?
I was a board member and rotated through most committees. It was a great experience to try my hands on many different aspects of CognAC. It has definitely helped improve my organizational skills, especially if the committee was formed by multiple study groups with people I did not know, like the Gala committee.
2. What is your most ‘typical CognAC’ memory?
It's too hard to pick just one. Maybe the most typical memory I have is how active a group it was. Reflecting on it today, the amount and size of events that were organized by what was, at the time, such a small study programme really amazes me. The event-packed study trips to Sweden and Scotland, the symposia, the autumn walks with pancakes, the movie marathon nights, to me are all typical CognAC.
3. What is/are the most valuable thing(s) you learned in university, outside from course material?
To not stress too much. There was always something coming up, either a deadline from a course or an event for which I still had to do one thing or other. Several friends helped me see how it is not necessary to stress out about each one of them. I definitely also picked up some organizational skills to help with that.
About your career
4. What is it exactly that you do now on a daily basis?
I am a software engineer for YouTube, working with teams focused on understanding content reuse. My daily work is roughly 40% meetings, 40% writing SQL and a surprisingly low fraction of "real programming".
5. What skills are most valuable in order to perform well in your workfield? What would you advise students to acquire these skills?
Using your time effectively. The number of open projects and ideas that need exploration are always outweighing the time that is available to you. The most valuable skills are to prioritize and to "fail quickly": find a way to quickly show that some idea will not work so you can move on to the next. It is something that mostly comes with experience. I read the previous entries in this blog and want to support Joris' reading suggestion of Getting Things Done (GTD) by David Allen. The book is a bit lengthier than it needs to be, but to me has been very helpful.
A more field-specific skill is the ability to quickly embrace a new programming / scripting / config language (or library, or other generic technology). Don't get attached to one set-up or to avoiding language X "because it is ugly". Companies (and different projects within a company) use many different technologies and this should not hold you back. Try to get a significant amount of experience in a large variety of programming languages, by taking courses on these languages or by purposefully choosing a language you are unfamiliar with when a programming project gives you that opportunity.
6. Are there any things you wish you’d have done in your time as a student?
I did an Erasmus exchange for 6 months in the last year of my Masters. Extending my study experience beyond one university/country was a great experience and in hindsight I should have looked for these opportunities earlier on. Living a month abroad or taking a (summer) internship at a company are things that I now would recommend my past self to do a lot more of. Not only to learn more about other cultures, but also to better understand the many opportunities that are out there.
Dealing with problems
7. What is the hardest thing about your current job? How are you dealing with that?
There can be a few too many chefs in the kitchen. I have many intelligent colleagues with different and valid thoughts about the projects I work on, meaning that often the way forward is unclear. One way that helps to deal with this is to write things down and get everyone to agree on what is written there, rather than having many meetings and discussions where a lot is said but very little is decided.
8. How did you come to your current job? Did you take steps in between? Could you consider any of these steps as ‘failures’, and if so, in what sense?
After my Bachelor in AI, I switched to a Master in Computer Science (with a focus on Computer Security). I made this switch because, at the time, I was not confident that the Master programme for AI was matching my interest or greatly help my job opportunities in the future. I may have been wrong about that, but I still really enjoyed the Computer Security specialization track I took instead. The switch did require me to take additional courses (not having a BSc in Computer Science), adding about a year to my studies
The Erasmus exchange during the last year of my Master provided the opportunity to apply for a 5-year PhD position at Chalmers University, in Sweden. I got the position and while I very much enjoyed my time there, after about 3 years I was seriously questioning whether this academic world was for me. I was able to do a summer internship at Google in the US during the 4th year of my PhD, which helped me decide to not pursue a postdoc position. Instead, I was able to convert my internship into a permanent position.
Considering the job I hold today, I definitely could have taken a much shorter path to get here. I don't work with Computer Security and the PhD title is kind of irrelevant, not being in academia. Still, to me it seems every career step somewhat naturally followed the previous one. It was also good to have this time to find out what I want. I'm more than fine with having taken the longer road.
Advice, to close off
9. What is the best advice you ever got? Why did it turn out to be so useful?
Honestly, I don't know what to answer here. I am sure I have been given lots of advice by fellow students, friends, mentors, colleagues and motivational speakers but I do not recall a single advice that was life-changingly useful. If I have to pick one from recent memory, maybe "take a day off". Extending your weekend by just a single day to increase the disconnect between work and life can do a lot of good for your stress levels. A luxury you likely don't have as a student, but something to remember for the future perhaps.
10. Do you have any good advice for ambitious students that would like to go into your workfield? And what kind of advice should they ignore?
If you are ambitious to be an engineer at a software company, do a lot of engineering work. Play around a lot. Run your own server. Write your own programming language and compiler. Take part in programming competitions, ideally in teams. Take a lot of internships. Teaching a field-related course will help you understand things so much better than taking the course.
Don't ignore any advice that you get. Most people who give you advice share with you something that they believe really works for them. So listen to every advice you get and see how it might fit into your life.
Maya Sappelli
With degrees in both AI and Linguistics under her belt, Maya Sappelli is currently active at the HAN as an applied Data Science researcher, mainly working with textual input. She once even wrote a name generator script based on character frequency in baby names, as she couldn’t decide on the name of her second (and currently youngest) child (the final name was not generated by this script, however ;) ). Found out more about Maya and her projects in her interview!
University and CognAC
1. Were you involved within CognAC? If so, how? What did you learn from that experience?
Yes in my first year I was involved in the activities committee. I mostly learned that it was easy to involve already active students, but hard to get the other students involved. A year later I started as secretary in InTenS (my second study association) and then I became less active in CognAC.
2. What is your most ‘typical CognAC’ memory?
My most typical CognAC memory is of course the TK, that was such a nice place to talk to other students and felt so cozy with the old mac computer there. Of course back then there were a lot less students.
3. What is/are the most valuable thing(s) you learned in university, outside from course material?
I was doing 2 course programs (AI and Linguistics) at the same time, which really upped my time-management skills. In a sense doing 2 programs also taught me to relax - I did not have to do everything perfectly - enjoying the things I was learning was good enough.
About your career
4. What is it exactly that you do now on a daily basis?
I am a researcher on Applied Data Science at the HAN (Hogeschool Arnhem Nijmegen). I started there a year ago and it is a really nice environment where I get to talk to a lot of people with various backgrounds on data and machine learning related topics.
5. What skills are most valuable in order to perform well in your workfield? What would you advise students to acquire these skills?
An important thing is to listen well to what people are saying and translate these things to data science or AI problems. Typically, I work with people that come from quite different backgrounds and domains who have interesting problems and data that can benefit from a technical solution. I have benefited a lot from the psychology courses but also the linguistics courses that I had. They help me view things from different perspectives and make it easier to connect and communicate to people with a different background so that we have a good mutual understanding.
6. Are there any things you wish you’d have done in your time as a student?
No not really, my student life was pretty packed with the things I wanted to do.
Dealing with problems
7. What is the hardest thing about your current job? How are you dealing with that?
Since I have only been working at this place for a short year, I am still getting to know a lot of people and getting to know the structure of the organization. Although teaching is not in my current job description, it will probably become part of it next year, because we are starting a master's program in Applied Data Science. I have given some lectures in the past, but being involved in creating a complete master's program from scratch is something else, which is why I will work on developing my didactic skills. I am looking forward to that.
Also, I consider myself a text miner - I usually work with textual data, typically unstructured free text. In this job I am more of a data science generalist, so I need to read up on some techniques for different data types as well.
8. How did you come to your current job? Did you take steps in between? Could you consider any of these steps as ‘failures’, and if so, in what sense?
I have always worked in research positions outside the university. First at TNO for 7 years, where I also did my PhD. I gained a lot of experience there in such a wide variety of activities (project management, acquisition, various text mining projects in different domains such as health and dark web).
At some point I wanted to get a job closer to home. I found a nice research project at FD Mediagroup on summarizing news articles, which was a really valuable experience since we were working on production systems, which was different from the proof of concepts I worked on at TNO.
Then - via my network - I found this job at the HAN - even closer by home. The nice thing about the HAN is that they do applied research: it is more focused at helping organizations with their current problems - you really work towards solutions that will be implemented in an organization in the short term.
Advice, to close off
9. What is the best advice you ever got? Why did it turn out to be so useful?
'Do your internship at a company'. When I was looking for an internship for my master it would have been easy to do one at the university - that was the safe choice. But because of this advice I did my internship at Philips which broadened my network, and I learned what it was like to work outside of academia.
That gave me a lot of insights about what I liked and didn't like, and helped me make a decision for the next step in my career.
10. Do you have any good advice for ambitious students that would like to go into your workfield? And what kind of advice should they ignore?
My advice would be to reach out to people already working in the field if you like what they are doing. Don't be afraid to ask for help or advice. In practice I think it is really helpful for a data scientist to have a portfolio or some open source work so that you can show your skills. That way you can showcase your strengths and what you find interesting.
Editors note: you can read more on Maya's projects on her personal website, https://msappelli.github.io/.