Avatar

Hello! I am Loïc Bachelot

A data scientist passionate about open science and earth system data!

I hold a Master’s degree in artificial intelligence and big data from the University of Cergy-Paris. My journey began in the airline industry with Air France-KLM, and later shifted toward ocean and earth sciences. At IFREMER, I contributed to the EU Blue-Cloud project, deploying deep learning models for ocean profile reconstruction and supporting marine bioacoustics research. Currently, I work at the University of Oregon as a Data Scientist with the Cascadia Region Earthquake Hazards Center (CRESCENT), where I lead cloud-native data platform development while continuing research in geophysical machine learning.

Check out my most recent work!

My resume

Some of my projects!

GNSS GNN Visualization

GNSS Time Series Denoising

First-author publication on applying Graph Neural Networks to denoise daily GNSS time series in the Cascadia region. This work bridges geodetic signal analysis and modern ML techniques, with an emphasis on open, reproducible science.

Podcast thumbnail

Podcast: ML for Earth Sciences

I was recently invited to discuss my GNSS denoising paper and my broader approach to applying machine learning in earth sciences. We covered reproducibility, open science, and the future of geophysical ML.

CRESCENT Platforms

As part of my work at the University of Oregon’s Cascadia Region Earthquake Hazards Center (CRESCENT), we develop interactive open-source visualization platforms: CVM (velocity models) and CFM (fault models), all built on AWS to support open-access, FAIR-aligned geoscience tools.

OSnet-GulfStream

This project was part of my job at IFREMER and resulted in the publication of a paper: Four-dimensional temperature, salinity and mixed-layer depth in the Gulf Stream, reconstructed from remote-sensing and in situ observations with neural networks.

Puzzle solving

Adversarial learning model applied to puzzle solving. This project was my research project for the last year of my Masters. It is part of a bigger project including a MCTS method (Johan's work) to solve puzzle.

Meteorite dashboard

I developed this interactive dashboard as a personal project to learn and improve in data visualization. I used meteorite data because we all would love to find one! Go play with the dashboard!

Polynyas

Exploring ice, ocean, and/or atmosphere interactions at persistent polynyas. This project was part of the ICESat-2 Hackweek 2022, working with the ICESat-2 data. I was part of the Polynyas team, studying these persistent water holes in the ice in Antartica.

Tympanogram tool

I developed this tympanogram creator tool in collaboration with Dr. David Brown, and Dr. Cassidy Bachelot from Pacific University School of Audiology. This tool is useful for students to learn the different values measured during a tympanogram exam. Go try it here!

Acoustic Sound Scattering Layers

Acoustic data contains information about organisms in the scattering layers. The project aims to build tools for classifying and interpreting this data. It was part of the 2021 Ocean Hackweek. I later continued working on this project with Muriel Dunn as part of her PhD. (paper coming soon)

Some of my other projects can be found directly on my github!

It includes school work like a cool robotic project of object tracking on Arduino, some Kaggle competition notebooks, spaciotemporal kriging in R, event detection and rating from tweets and more!

Find me on ...