I really value efficiency. I believe that if we can make things efficient, from a holistic point of view, it frees up time and resources for more important things in life and society.
That's why I focus on the implementation of AI. There's so much talk, hype, and proof of concepts in AI right now. There's also a lot of potential. But potential, in and of itself, isn't going to create efficiency. That's why I always push for actual use cases of AI.
I did my bachelors in economics. I liked statistics, and, since they're pretty related, I started doing some machine learning.
This led me to change my academic focus. Instead of economics, I did my master's in data science. I enjoyed modelling, especially the Explainable AI side of it. That's also what I did my master's thesis on. And it's still my field of expertise.
After my studies, I went into AI consultancy, starting out on the data science side (with a bit of data engineering too). I think it's important to get the full spectrum, especially if you want to go for implementation and impact. But, as it was consultancy, I also moved into project and account management. I think I'm stronger when I'm doing all of these things together and making sure that the tech lands.
I worked on AI projects for over a dozen companies, from big corporates to startups, before joining Kickstart AI.
I'm an AI Project Lead at Kickstart AI. I work with our partner organisations on projects. My goal is to ensure that these projects and our partners succeed.
When we start a project, I am very involved in defining the goal, or challenge we want to solve, with a partner. Once that's clear, and we need to start developing, I shift more to account management. My colleagues then do most of the development and day-to-day project management. And they report to me.
I significantly influenced our project approach at KAI. The end-to-end methodology we use was my proposition.
I’m really proud of a project we're doing with Ampere, which focuses on optimising logistics by predicting parcel volumes and weights. We've made big improvements. And we'll share the results publicly soon. This will translate into less kilometres driven, so less emissions, and more efficient operations overall.
This project shows how AI can bring tangible benefits to a company's core operations. We've gone from inception to setting up an AI that creates significant business value in their daily operations in less than a year.
I also enjoy the project we're working on with NS, related to improving their safety processes, as it's very technically challenging. We’re using state-of-the-art models in areas like computer vision - models that literally came out a month ago. I don't think many people can say they are using these models in their actual work, outside of experiments. So that’s really cool.
I believe Explainable AI (XAI) is crucial for AI adoption. When you're implementing an AI model in business operations, it's kind of like introducing a new employee — a very strange employee. And it can be difficult for people to understand them. It's not that people are distrusting of the AI. But it often feels foreign. And they often don't have a feeling about the implications.
Explainable AI is like having a translator between people and AI. It’s important to be able to give context, and explain what’s happening in the background.
Every successful project I’ve worked on has included some sort of Explainable AI. I believe it’s key for implementation.
I use ChatGPT. I’m very careful, though, not to feed it sensitive information. I find that very important.
I'm currently also doing an experiment with AI assistants. So I've got some apps running for that. I see the potential of using them as sounding boards and experts to discuss things with for my work. Will they stay? I don't know, but it's intriguing to explore.
I don't follow any blogs or web sources for that. And that’s because of my colleagues at Kickstart AI. They're just so on top of everything! They use it for projects every day of course. So honestly, just by having lunch at the office, I hear about all the latest concepts and trends. Which is great!
Generally speaking, I think it will feel very smooth and seamless. It will be an integral part of most people’s everyday work.
I can imagine myself in 5-10 years, having a conversation with a new employee who just finished college, and telling them anecdotes about my early days of consulting. They'll be like, “What? You wrote every email yourself!?”, or “You actually did the designs of the PowerPoints?”
I think we’ll realise that, wow, the way we work is very different because of AI.
The internal drive I have for AI adoption is really in line with what Kickstart wants to do. So it's a great match there. I also love the vibe. We've got a shared mission and a lot of really smart people here – so it’s a stimulating environment to be in.
I think it’s important to take a well-rounded approach. If you want to be on the AI implementation side, I don’t believe there is such a thing as just being just a techie or just business-oriented. So I think a well-rounded education is key. So sure, do math, but also do social sciences.
💡 Want to learn more about Kickstart AI and our mission to accelerate AI adoption in the Netherlands? Check our website or connect with Sophia on LinkedIn!
