Case study

Machine learning engineer and co-founder of Circular — Jeremy Bertincourt

Having got to grips with the basics of machine learning engineering in his own time, Jeremy went on to take a nanodegree through Udacity, before founding his own company

How did you get your job?

I obtained a Masters degree in engineering, which covered electronics, computer science, maths, physics and finance and included a major in embedded systems/artificial intelligence.

My first job was working as a software engineer in a big French company, but I found this environment very slow and not challenging enough for me. I moved on to a different job in Germany, where I worked on autonomous vehicles.

I then discovered an interest in image recognition and the field of machine learning. I studied it in my free time, discovering algorithms, working on several projects while studying a nanodegree with Udacity, and participating in a Kaggle competition.

After a year of working as a machine learning engineer in a start-up company in London, I decided to switch to a small consulting company, providing data analytics and machine learning solutions to big companies. It gave me a better insight into big data and how to efficiently manage huge data sets.

I then returned to Paris to look for a job in a start-up. After a few weeks, I was contacted on AngelList by other company founders who were looking for someone who specialised in this field, which is when we co-founded Circular.

How relevant is your degree to your job?

An engineering degree teaches you how to think scientifically - very important in machine learning. It also gave me the fundamentals in programming, maths and physics, which made my machine learning nanodegree easier. It taught me about micro-controllers, which is helpful when working with my colleagues who specialise in hardware, as it enables me to understand the constraints they face.

My job involves building the back-end on the cloud, which requires software/data engineering skills (taught in my engineering degree, but not in the machine learning degree) as web and OS languages might need to be used.

How would you describe your working day?

I work from home or from a café in Paris to build new functionalities for our product using machine learning. I also take part in regular discussions with all the founders, about new advancements in the company.

How has your role developed?

The combination of study and work experience I have built up gave me the necessary knowledge, skills and expertise to set up my own company.

At Circular we've been working on a connected ring designed to solve the problem of exhaustion in cities. It increases your energy by improving your sleep and pushes you to do more daily activities with the help of a community and AI. The ring is still under development but should be ready to be rolled out commercially soon.

I think running my own business is what I was made for and might spend the rest of my life creating companies.

What do you enjoy about your job?

I love machine learning, building a product instead of working on pure data; the freedom and flexibility I get from being a co-founder, and the creativity of product creation.

What are the challenges?

When building a new company, you need to be ready to take risks to achieve your goals.

You won't get a salary at the very beginning and you need to invest some money to build the capital. You need to understand who your customers are and what problem you are solving. You also need to take ownership of what you build. If the server is down during the weekend, you need to fix it during the weekend, not on Monday at 9am.

Any advice for someone who wants to get in to machine learning?

To get into machine learning, you can either do a Masters or PhD (preferred) in this field, or start in software and learn machine learning through an online platform.

Base your degree choice on its reputation and quality, but also most importantly on what people have become with it.

You could ask an employer if they will finance an online course in this field. It doesn't guarantee you will apply models in your day-to-day job, but it will give you the opportunity to use your software skills in machine learning degree projects. You could then find work in, or set up your own start-up company.

Find out more