Diogo Netto
Software Engineer at RelationalAI
Old Toronto, Ontario, Canada
Summary
🤩
Rockstar🎓
Top SchoolDiogo Netto is a software engineer specializing in programming language runtimes, performance engineering, and low-level systems with six years of professional experience. Based in Toronto and educated at MIT (MEng and dual bachelor's), he has made production contributions to Julia’s GC and allocator—shipping parallel and concurrent sweeping, a scalable memory pool refactor, and a coroutine sampling profiler merged into Julia releases. He also contributes to scientific ML tooling, integrating and validating advanced ODE solvers and a continuous normalizing flow layer in the SciML ecosystem. Diogo combines deep systems-level optimization with numeric and ML-focused engineering, comfortable moving between garbage collector internals and high-performance numerical algorithms.
6 years of coding experience
Master of Engineering - MEng, Computer Science, Master of Engineering - MEng, Computer Science at Massachusetts Institute of Technology
Github Skills (20)
algorithm10
algorithms10
ziggy10
ode10
ordinary-differential-equations10
mathematics10
machine-learning10
mathics10
math10
maths10
differential-equations10
zig10
adaption10
numerical-methods10
neural10
Programming languages (3)
Github contributions (5)
SciML/DiffEqFlux.jl
Jun 2020 - Aug 2020
Pre-built implicit layer architectures with O(1) backprop, GPUs, and stiff+non-stiff DE solvers, demonstrating scientific machine learning (SciML) and physics-informed machine learning methods
Role in this project:
ML Engineer
ML EngineerContributions:121 commits, 21 PRs, 39 comments in 2 months
Contributions summary:Diogo primarily contributed to the implementation and testing of a continuous normalizing flow layer (CNF) called FFJORD, designed for density estimation within a scientific machine learning context. They developed the FFJORD layer, incorporated support for multivariate distributions, and added comprehensive tests using the Beta and Normal distributions. Furthermore, the user addressed documentation and code structure issues, demonstrating proficiency in applying CNFs to real-world problems.
scientific-machine-learningpartial-differential-equationsdifferentialodescientific-ml
SciML/OrdinaryDiffEq.jl
Mar 2020 - Jun 2020
High performance ordinary differential equation (ODE) and differential-algebraic equation (DAE) solvers, including neural ordinary differential equations (neural ODEs) and scientific machine learning (SciML)
Role in this project:Back-end Developer
Back-end DeveloperContributions:44 commits, 17 PRs, 14 comments in 2 months
Contributions summary:Diogo implemented and refined the KYKSSPRK42 algorithm, a specific numerical method for solving ordinary differential equations. They modified the `ssprk_perform_step.jl`, `ssprk_caches.jl`, `alg_utils.jl`, and `OrdinaryDiffEq.jl` files to incorporate and test the algorithm. Furthermore, the user added convergence tests and corrected coefficient issues. The user's work focused on integrating, optimizing, and validating a new numerical method within the SciML ecosystem.
adaptiveodesscientific-machine-learningdifferential-algebraicdifferential
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.