Ben Frederickson
Principal Engineer at NVIDIA
Vancouver, British Columbia, Canada
Summary
🤩
Rockstar🎓
Top SchoolBen Frederickson is a Principal Engineer with 12 years of experience building high-performance, GPU-accelerated data and recommender systems at NVIDIA and earlier roles at Amazon and Flipboard. He combines systems-level engineering—cross-platform profiling, build and CI improvements, and containerization—with applied ML work such as GPU-driven nearest neighbors, collaborative filtering, and feature engineering for terabyte-scale tabular data. A prolific open-source contributor, he’s improved projects like NVTabular, cuDF/cuML/RAFT and NMSLIB by adding pybind11 bindings, Parquet streaming, approximate NN optimizations, and production-ready operators. He frequently bridges back-end, DevOps, and ML responsibilities, shipping production infrastructure and unit-tested numerical code. Based in Vancouver with a background in cognitive science and engineering from Simon Fraser, he brings a rare mix of low-level systems rigor and practical recommender/ML product experience.
12 years of coding experience
16 years of employment as a software developer
BA Cognitive Science, BA Cognitive Science at Simon Fraser University
Stackoverflow
161reputation
13kreached
10answers
0questions
Github Skills (75)
vector-search10
python10
nearest-neighbors10
testing10
windows10
bash10
feature-engineering10
gruntjs10
ruby10
gpu10
cython10
pybind1110
parquet10
cpp10
data-pipeline10
Programming languages (13)
Github contributions (5)
benfred/venn.js
May 2014 - Nov 2018
Area proportional Venn and Euler diagrams in JavaScript
Role in this project:
Full-stack Developer
Full-stack DeveloperContributions:9 releases, 71 commits, 14 PRs in 4 years 7 months
Contributions summary:Ben's primary contribution involved significant source code reorganization of the JavaScript files. This included moving all JavaScript code into a `src/` directory and setting up Grunt for concatenation, minification, and testing. This restructuring aimed to improve code organization and potentially enable exporting the code for use in Node.js environments. This work involved modifying a core layout file, suggesting a central role in the project.
javascriptvenn-diagramd3euler-diagram
benfred/implicit
Apr 2016 - Dec 2022
Fast Python Collaborative Filtering for Implicit Feedback Datasets
Role in this project:Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:1 release, 15 reviews, 306 commits in 6 years 8 months
Contributions summary:Ben's contributions primarily focused on the development and optimization of a collaborative filtering algorithm for implicit feedback datasets. They implemented and improved the Bayesian Personalized Ranking (BPR) model, added functionalities like GPU accelerated inference methods and incorporated code for running evaluations on the GPU. The user also refactored existing code for the Approximate Alternating Least Squares (ALS) models with the integration of a high-performance top-k ranking function for item recommendations, improving the system's efficiency.
collaborative-filteringpythonmachine-learningmatrix-factorizationrecommender-system
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.