Claire-Isabelle Carlier

MADS 2021

Claire-Isabelle Carlier was a member of the first cohort of the University of Michigan School of Information’s fully online Master of Applied Data Science degree program. She came to the program with a background in translation and multicultural management, and a budding interest in data analytics. 

At UMSI, Carlier found the support and opportunities she needed to make a successful transition to “full-fledged data scientist.” She graduated in 2021 and is now a senior data scientist at Engineered Intelligence, a startup providing software and modelling services to utilities, where she liaises between subject matter experts and the development team to build the right models for their clients.

How did you discover the MADS program at UMSI and enter the data science field?

My career path took a sharp turn a few years ago. I became a business analyst and later an enterprise architect at a renewable energy company. While there, I became interested in data analytics. That's when I discovered the School of Information's Coursera online courses, such as "Python for Applied Data Science." I completed these courses and signed up to be kept in the loop. Then I heard about a master's program opening up. I saw it as an opportunity to formalize my transition to a data scientist role, having no prior academic background in the field.

During the program, I not only validated my previous work but also consolidated my base knowledge in statistics and programming to become a full-fledged data scientist and now be able to claim, “Here, I have a degree to prove that.” Before graduation, I got a data scientist role, and then a senior data scientist role, at Canada Post where I worked for four years. Most recently I transitioned to a new role as a data scientist at a startup in the field of electricity grids

The journey required hard work, but I'm grateful for the resources available online and the support of the UMSI community. It played a significant role in my successful career change. I'm definitely grateful for that.

What do you do in your current position?

In my current role as a senior data scientist, I am part of a team of consultant data scientists available for advanced analytics projects within our organization. These projects cover a wide range of challenges, often requiring complex data manipulation, programming and statistical knowledge. 

Often, new people on board expect there will be crazy models. That's not true. Not all organizations are mature enough to have machine learning and deep learning projects. We focus on descriptive analysis and forecasting. An example would be forecasting how much our volume is going to evolve, then how to adjust to that from a plant perspective, a delivery perspective, a training perspective, and also investment-wise. If we know that our volumes will increase in this province of Canada, we want to prepare for that. But we don't want to prepare for that everywhere, right? We need only adjust service at a localized point. The diversity of these projects can be challenging, given the need for in-depth domain knowledge. So it’s always a collaborative effort with field experts and client teams to ensure accuracy and validity.

I had a really fun project recently where we did a scenario analysis on our equipment: Should we change the size of our parcel mailboxes? We had to keep upgrading costs down, but if items don't fit in the mailbox and people have to go to the post office to retrieve them, it'd be more work for us. 

The first six months when you start in our team are really challenging. The knowledge that you have to gain quickly is immense. We have a wealth of data sources, and while we may not necessarily document everything, we encourage juniors to reach out to senior data scientists for guidance. We always pair juniors with senior data scientists, so we offset a bit of that load at the beginning. It takes time to get comfortable with data, concepts and business rules.

Where did you work before this? How was that experience compared to the experience at your current company?

Before my current role, I was an enterprise architect in a team setting strategic guidelines for this big multinational operation. We were establishing best practices for data capture and storage. We dealt with very specific challenges, such as some of our sites being very remote and not always having wifi connection. In that case, what's important to capture, and what's not? What can we do with such data?

I got to work on machine learning pilots, primarily in collaboration with wind farm teams as they were the most mature when it came to data. We used computer vision to automate the analysis of drone-captured images during farm inspections, to identify cracks or other defects on the turbines that we needed to tackle. We were required to announce how much energy we could produce in a day, so there was a lot of data gathering. Working in the energy sector was important to me as it allowed me to contribute to climate change mitigation.

In my role, I handled a lot of pilot projects, which, while enjoyable, never went into production. I talked to my boss at the time about going into data science full-time, but they were not ready to have this role full-time, and that's why I ended up switching to a different organization.

At Canada Post, I try my best to be assigned to projects that reduce carbon footprint or optimize delivery to reduce gas consumption. That's what makes the work I like, contributing to a good cause. 

Do you have a career update you would like to share, such as a past or future project that you're excited about?

I was very happy I got to speak at a Big Data and AI conference in Toronto. I got to talk about projects that I was doing at work and got a lot of interest from the audience. It was the first in-person conference I attended in years since COVID. Presenting to a large crowd at a big venue was very rewarding.

The other thing is our collaboration with municipalities and universities in Canada in 2023, to study congestion and its impact on emissions and climate change. We have ongoing projects with Vancouver and Montreal, as well as universities. I love collaborating with universities because you gain such a unique, specialized set of knowledge. I’m looking forward to an opportunity to learn, working with a broader network of professionals outside my regular team, and the positive impact this work can have.

What does a day in your life look like? How do you balance work and life? 

I'm not very good at that. My days are quite intense.

It's a big challenge to be a manager and a technical resource at the same time. Because as a technical person, you always want to do the work on the project, but at the same time, you have to let the junior people learn, try, make errors, then correct them. I'm learning to strike a balance. 

Much of my role is about client relationships. I have many client meetings, where we discuss project scopes and updates and adjust timelines because we never have enough time to do projects. I also mentor junior data scientists and allocate them to various projects.

Within each team, we plan our methodologies, distribute tasks and collaborate on advanced work. I believe in sharing knowledge because no one knows everything in this field. There's so many techniques out there. Of course, we also do a lot of programming, preparing slides and reports on outcomes.

I often spend time coaching junior data scientists, which I find enjoyable. It involves a lot of human interaction, which sets our team apart from those who prefer to work in isolation. I sometimes realize at the end of the day that I've spent more time talking and helping people than doing technical work. So, I make an effort to allocate time after hours to focus on my tasks. Balancing work and life in this role is challenging, but I'm determined to improve.

In this kind of data science and consulting setup, you have to push back on unrealistic deadlines and prioritize important projects. Sometimes, it's important to say, “No, we can't do this. We have to push the deadline to December.” We will work on it as soon as we can, but not right away. I've been successful in reevaluating project priorities this past year, and I want to maintain this approach in the future.

Do you find yourself applying the skills that you learned at UMSI in your career?

Yes. The thing I liked is that the MADS program has a project-based approach. There were two capstones and one milestone, both self-directed. You have to define your own scope by your idea, then make it happen in the end. This approach mirrors real-life projects where clients may have a vague idea of what they want, and it's your job to help define the project scope, objectives and value while you gather the necessary data.

I know a lot of people may not consider communication skills to be important, but I had an entire course where I learned how to do a good presentation in PowerPoint. This ends up to be a big part of what I do at work. Yes, I do a lot of programming, but everything else is communications, and my audience needs to listen to me and know exactly where they can take action, make a decision, or sign off on an investment. If I don't present the information accurately and concisely, nothing will ever happen. 

As data scientists, we draw a conclusion from the analysis we've done, and present this to someone else in a way they can understand. An executive may not like crazy statistics, and prefer to get to the point. They don't care how we got to that conclusion. 

It’s worth noting that what we deliver is solutions, not techniques. While we can implement novel, complex models, it's equally important to recognize when simpler or existing approaches can achieve the same results more efficiently. Overly complicated models are often challenging to maintain and are not necessary for all tasks. So we have to learn where to use the fancy techniques, and where not to use them.

Was there a resource at UMSI that was particularly helpful for you?

I had the opportunity to join a project in collaboration with the School of Public Health, building a COVID-19 dashboard reporting cases and tests at the region level in the state of Michigan. What made this experience unique was that it put my skills to the test in a professional setting.

Working on this project was a game-changer because I could participate in a project that I get to present to a state decision-maker, collaborate on tools like Github and VS Code, and move beyond individual work on Jupyter notebooks. The urgency of the pandemic demanded that we worked very hard to build a critical dashboard quickly. This was a learning experience that pushed me, putting my skills from one to five immediately.

When you have something in production, things never stop just because you've done the first step right. You continue, add, improve, optimize, and most importantly, maintain it. I only stopped working on this project at the end of 2023 when it was announced the dashboard would be decommissioned, so I followed and supported this project for over three years. I learn new things every day because of that collaborative effort and professional setting. 

While some students can engage in research projects during their master's, not everyone gets the chance to work on real-life projects with such impactful outcomes. It was an opportunity of a lifetime that gave me a preview of the challenges and rewards of working on data projects.

Did you find it challenging to break into a tech career? 

It's a good question. I always believe that if you demonstrate your abilities, people will let you in. If you show that you can do the work and do it well, you earn respect for that. I am lucky to say I never felt unwelcome or discriminated against since I switched to a tech career. 

I believed very strongly in the fact that I was putting many hours into learning new skills, and the learning paid off. I never think about the difficulty of getting there. I think of time as an investment, and when you invest like this, know that it pays off eventually.

Most of my life, I have always gone where I initially did not belong. I'm an immigrant. I moved to Canada more than ten years ago, and it was the same process. It took me a few years to prepare to move, a few months to find a job, get to Canada, then get through paperwork. Switching careers was pretty much the same thing to me. It is a lot of work, but you have to put in the work to succeed. Maybe in five years, I'll switch careers again. Who knows! 

Try to frequently check in with yourself: Are you happy with where you are? If not, where do you want to go? How do you get there? I realize not every situation can be like that. To me, if you put in the work, you'll get there in the end.

Would you like to share one piece of advice for new graduates or current undergraduates trying to get their foot in the door in the information field?

The first few years are always challenging. 

Understand that you may not immediately work on your dream project. A lot of people I interview come in wanting to work on computer vision projects. Well, while I may not have a  computer vision project right away, I may have one in the next year or so. You have so many things to learn first, and you want to prove to your seniors that you have basic skills before being assigned more complex projects. 

Being passionate and showing that you can learn is very important. I don't like when people show up to interviews just looking for a 9-5 job, because I'm not going to enjoy working with that person. I want somebody who's passionate, interested in the work. If you're passionate about your work, it never feels like work and you learn so much faster.

After your first exposure, focus on areas where you genuinely see yourself fitting in. Aligning with the business domain and industry you're interested in can set you apart when presenting yourself to potential employers.

When people ask “Why would you like to work with us?” consider having an idea of a project you'd like to work on within that organization. It shows that you've thought about your fit within the company and are genuinely interested. Having a clear answer to this question can make a strong impression.

Ultimately, remember that you’re not after any job, but the right job for you.

Do you have recommendations for how recent graduates should navigate the tech workforce?

I understand the appeal, when you're a junior, to find your way into a fancy, big tech company. But you may find it more rewarding when you work in a mid-sized organization, because you have flexible rules, you get to work on and learn about several different things. When you don't know exactly what you like most and what you are better at, having that versatility helps a lot. In smaller companies, you may also more quickly move up to decision-making positions.

— Tuesday, July 23, 2024