Enjoying a nice seafood meal with fine Italian wine on the beautiful Italian coast in Camogli

Second Figure — A view of the Kasbah of the Udayas in Rabat, Morocco, overlooking the Bou Regreg river. I used to sometimes skip school to play football here and then enjoy a swim

Third Figure — Semi de Paris, one of the biggest half-marathons, was my first race in 2021. I finished with a time of 1h47min

Fourth Figure — LondonT100 2024 Triathlon finisher with massive quad pain: nothing special, just doing a sprint triathlon with a city bike that weighs 3 to 4 times more than a racing bike with bad aerodynamics. The reason? I couldn’t find a rental racing bike, and honestly, I was a bit worried about getting injured using one for the first time in a triathlon. At least the city bikes felt 100% safe since I used them every day to get to work

Learning piano

Fifth Figure — Celebrating my Chess learning curve

Welcome to My Blog

Hi, I’m Amine and I use this blog to share ideas, mainly related to my field of expertise (data science, statistics, and ML), but also on other topics of interest such as finance, physics, philosophy, chess, sports, nutrition, and music.

Background

I graduated as an electrical engineer specializing in robotics and ML in Morocco. I was heavily involved in robotics and 3D printing, building robots (check my post about Sarah, the humanoid robot I built).
Back then, I couldn’t imagine working with just numbers, without any hardware involved. I must admit, it’s been a helpful shift for my lifestyle.
I no longer have a bunch of wires coming out of my laptop, motors plugged into it, etc. Now, my entire lab is inside my laptop, which makes traveling (something I do quite often) and working much easier.

Career

Throughout my career, I’ve worked on a diverse set of problems:

Traffic modeling, parking, and city dynamics at Datategy – I worked on these topics early on in my career, including image processing for parking spot detection and parking affluence forecasting at a French startup in Paris. Here, I learned how to handle real-world data and understand the dynamics of urban environments, which was key to developing smart city solutions.
Household energy consumption modeling at Engie France B2C – Our goal was to help customers monitor and optimize their energy usage. This involved heavy forecasting work, from long-term (yearly forecasts) to short-term (hourly forecasts). I developed sophisticated models, including attention-based, seq-to-seq encoder-decoders, while also seeing firsthand how simple tools like Fourier transforms can be incredibly effective in the right setup. This work covered many areas, such as uncertainty modeling, and introduced me to various key forecasting tools. Additionally, I gained significant experience in A/B testing and hypothesis testing, as online experiments were crucial to understanding customer behavior in this role.
Programmatic ads at Amazon UK – This is where I was introduced to the fascinating world of programmatic advertising. Imagine matching millions of ads to hundreds of millions of advertising opportunities in real-time. From the moment you click on a webpage with an ad slot to the moment the page fully loads (around 50 milliseconds), a complex pipeline runs including a real-time auction to determine which ad to show you. This decision is based on various factors like your behavioral data (cookies), the context of the webpage you’re browsing, and the likelihood that you’ll interact with the ad. The goal is for the advertising company to maximize its revenue. The field is very NLP-heavy, but also involves sound and image processing (especially since Twitch is part of Amazon). LLMs (Large Language Models) play a significant role across the pipeline, helping to process text and context in real-time.

Throughout these projects, I developed my business acumen. Whenever I start a project, I ensure I have a clear understanding of the demand-supply dynamics, identifying the core business problem and working backward from it. Many scientists, especially those passionate about the science itself, tend to get caught up in the beauty of the tools and theories, sometimes diverging from the real business needs. This is one of the key differences between working in a business environment versus working in a lab.

Interests

I grew up in Rabat, which is probably why I have such a strong connection to the sea. I spent a lot of my childhood playing in Rabat’s lovely streets.

Sports play a big part in my life. During COVID, I got into endurance sports, which was a paradigm shift for me. I moved from playing football, which is social and offers immediate rewards (scoring a goal), to long runs, sometimes 2 hours or more, where the reward and feedback loop is much more complex (the long-term benefits endurance sports have on the body). If you’re lucky, you experience a fast reward with an endorphin release during the runner’s high, but often, it’s the long-term gains that count.

I love learning new things and setting challenges for myself. I started playing chess in 2021, piano in 2023, and I’m still struggling with Spanish. I’m also very interested in finance.

I’m fascinated by intelligence, wherever it exists. I believe intelligence is spread everywhere, but the curse of expertise—focusing on a small subset of problems for the rest of your life—can sometimes prevent you from interacting with other forms of intelligence.
You’d be surprised by the things you can learn just by tracing the etymology of a random word like بلارج (pronounced “Bellarj”), which is how we refer to a stork in Morocco. The word has interesting roots, as it comes from the Greek word Pelargos, showing the historical interactions between the Greeks and Morocco.

Learning new things is my way of sampling from this vast space of intelligence.

Contact

If you think we can have an interesting discussion or collaborate on a nice project, feel free to reach out: LinkedIn.