Nico De Vos
AI & Accessibility Architect at Salesforce
San Jose, California, United States
Summary
🤩
Rockstar🎓
Top SchoolNico De Vos is an AI & Accessibility Architect based in San Jose with 13 years of experience building production-grade ML systems across accessibility, CRM, energy analytics, quantitative finance and earth sciences. He has led architecture and product vision on Salesforce’s Einstein Platform, shipping AutoML-powered features that democratize ML and designing reusable, scalable frameworks adopted across teams. Previously he architected long-running forecasting and real-time streaming services for millions of devices at AutoGrid, combining Python and Spark/Scala for high-throughput deployments. An active open-source contributor, he created a Python k-modes clustering implementation and improved Spark job grouping in Salesforce’s TransmogrifAI, reflecting a focus on practical, reliable tooling. Trained as a hydrology PhD, he blends rigorous scientific modeling with pragmatic engineering to turn complex data into operational insights.
13 years of coding experience
10 years of employment as a software developer
Doctor of Philosophy (PhD) Hydrology and Water Resources Science, Doctor of Philosophy (PhD) Hydrology and Water Resources Science at Delft University of Technology
Dutch, English
Github Skills (13)
machine-learning10
spark10
automated-machine-learning10
python10
numpy10
scala10
algorithms9
pipeline9
dsl9
feature-engineering9
transformers9
algorithm9
data-analysis8
Programming languages (8)
Github contributions (5)
nicodv/kmodes
Aug 2013 - Dec 2022
Python implementations of the k-modes and k-prototypes clustering algorithms, for clustering categorical data
Role in this project:
Back-end Developer
Back-end DeveloperContributions:10 releases, 7 reviews, 331 commits in 9 years 5 months
Contributions summary:Nico created the initial implementation of the k-modes clustering algorithm in Python, including the core logic for centroid calculation and cluster assignment. Subsequent commits updated and refined the algorithm's functionality, fixing bugs related to centroid initialization and cluster assignments. Further commits involved refactoring the code to improve clarity and efficiency and adding example code.
implementationspythonprincipal-component-analysisk-meansgaussian-mixture-models
salesforce/TransmogrifAI
Mar 2020 - Mar 2021
TransmogrifAI (pronounced trăns-mŏgˈrə-fī) is an AutoML library for building modular, reusable, strongly typed machine learning workflows on Apache Spark with minimal hand-tuning
Role in this project:Back-end Developer
Back-end DeveloperContributions:1 release, 32 reviews, 39 commits in 1 year
Contributions summary:Nico primarily focused on enhancing the core functionality of the TransmogrifAI library. Their contributions involved implementing and refining Spark job grouping, which is crucial for distinguishing steps in the machine learning workflow. They added job groups for various stages such as data reading/filtering, feature engineering, and cross-validation, improving the monitoring and management of these processes. Additionally, they updated and refactored existing code, introducing enums and logic to improve code structure and efficiency.
pythonsalesforcetransformationstransmogrifyestimators
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.