Hello, I'm Sixte

Things I do

I am a machine learning engineer at Panakeia AI, working on breast cancer diagnosis using and developing deep learning computer vision models.

Before that, I completed a masters of engineering in computing (AI and machine learning) at Imperial College London where I had the opportunity to conduct some research in adversarial machine learning. Aside from uni I used to work part-time as a junior developer at Blockchain in London which I found very exciting and useful to gain experience. It taught me a lot about Node and React as well as working in a team on software with more than 30 million users. Above all, it turned me into an investor and explorer of the crypto world, passionate about cryptocurrencies, digital assets and the underlying technologies. If you like blockchain technology or are simply curious you can play around with this football exchange DApp I built with React and Solidity on the Ropsten test network.

I also worked as a freelance developer besides my studies as well as on various personal projects including the analysis of biomarkers with machine learning for early cancer detection.

For a quick overview of what I've been up to since the 22nd December 1997 please read my CV.

Things I like to do

  • Write great code
  • Read books and news
  • Have good ideas
  • Travel

Projects of interest

As you're reading this I'm likely to be in the process of completing one of the following, if not all.

Visiting every country in the world

France, Spain, Portugal, Italy, Holy See, Switzerland, Germany, Belgium, England, Scotland, Austria, Hungary, Czech Republic, Poland, Sweden, Romania, Greece, Turkey, Jordan, Israël, South Africa, Puerto Rico (USA), Australia, Senegal, Oman, Thailand, Cambodia, Vietnam, Myanmar... I've visited 29 countries in 21 years and that is about 15% of what I plan to achieve in my lifetime excluding countries that do not exist yet.
In the years to come you may find me somewhere holding the skyscanner app in one hand and my thumb raised with the other to increase this score but more importantly discover new cultures, landscapes and lives.

One noise to fool them all

One noise to fool them all is my masters final year project (thesis) submitted as part of my curriculum at Imperial. In this project, I implemented the first ever query efficient attacks against deep convolutional networks for object detection and semantic image segmentation using procedural noise and bayesian optimisation. It was awarded a final grade of 90/100 and the Winton Capital Applied Undergraduate Project Computing Prize. It also gave me the chance to contribute to the paper Procedural Noise Adversarial Examples for Black-Box Attacks on Deep Convolutional Networks which was accepted at CCS 2019.

Etheroscope

Etheroscope was one of my software engineering group projects at university. With five other students from Imperial, we built an explorer for smart contracts on the ethereum blockchain. The aim was to provide in-depth insights about the evolution of the state of smart contracts.
At the end of our group project we will deliver the etheroscope to Alice.si, a company with social impact.
The technologies we used are Elixir and Phoenix on the backend and React on the frontend. Our open source code is available on Github.

Voix

Voix is an iOS messaging app that allows users to communicate seamlessly using Google speech API to transcribe all voice messages. It provides fast sending via voice messages while maintaining nice and easy receiving of messages.
Created in a webapps group project at Imperial College London.

Things you can do

If you want to reach out to me, just drop me an email at sixte@demaupeou.com

Or you can play Flappy Sixte, a game I made when I was in high school.

Alternatively you can also learn from my part of a wider research project about the Hopfield neural networks.

If you are still here, take some time to visit the website I made for my mom's paintings.