Jeremie Desgagne-bouchard
Science Advisor at Evovest
Montreal, Quebec, Canada
Summary
👤
Senior🎓
Top SchoolJeremie Desgagne-bouchard is a science advisor and modeling specialist with 10 years of experience bridging actuarial rigor and machine learning to solve pricing, cost analysis and portfolio optimization problems. He designs end-to-end ML frameworks—from data acquisition through strategy monitoring—and has tackled time series and uncertainty with bespoke optimized methods. His background includes R&D across text, spatial and telematics data and hands-on contributions to notable open-source projects like Flux.jl and Apache MXNet, where he improved RNN implementations and visualization tooling. Comfortable switching between technical depth and business pragmatism, Jeremie combines actuarial foundations with applied research experience at firms such as Intact, Element AI and Evovest. An unusual strength is his focus on problem identification and model mechanics, not just algorithm application, which helps turn complex analytics into actionable strategies.
10 years of coding experience
1 year of employment as a software developer
Bachelor’s Degree, Actuarial Science, Bachelor’s Degree, Actuarial Science at Université Laval
English, French
Github Skills (13)
neural-network10
mxnet10
machine-learning10
recurrent-neural-networks10
visualization10
deep-learning10
visualizations10
r10
gpu-acceleration9
cgan8
dcgan8
data-science8
cuda7
Programming languages (9)
Github contributions (5)
apache/mxnet
Jan 2017 - Feb 2019
Lightweight, Portable, Flexible Distributed/Mobile Deep Learning with Dynamic, Mutation-aware Dataflow Dep Scheduler; for Python, R, Julia, Scala, Go, Javascript and more
Role in this project:
Back-end Developer
Back-end DeveloperContributions:15 commits, 15 PRs, 151 comments in 2 years 1 month
Contributions summary:Jeremie primarily contributed to the R package within the MXNet repository, focusing on enhancing the visualization capabilities for computation graphs. Their work involved refactoring the `viz.graph` function, updating and improving the display of model graphs using libraries like DiagrammeR and visNetwork. Furthermore, the user added and refined examples and demos related to GAN training within the R environment.
pythonschedulerdataflowmutationdata-science
FluxML/Flux.jl
Oct 2020 - Mar 2021
Relax! Flux is the ML library that doesn't make you tensor
Role in this project:ML Engineer
ML EngineerContributions:8 reviews, 37 commits, 6 PRs in 4 months
Contributions summary:Jeremie primarily focused on developing and testing machine learning models within the Flux.jl library. Their contributions include implementing and testing RNN architectures, including vanilla RNNs, and custom recurrent layers, demonstrating a deep understanding of recurrent neural networks. The user also explored GPU acceleration and debugging potential issues with GPU execution within the models. Their work involved creating test cases and improving the stability and performance of the RNN implementations.
ml-librarythe-human-braindata-sciencedeep-learningneural-networks
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.