Amanda Dsouza
Staff Applied Research Scientist at Snorkel AI
Seattle, Washington, United States
Summary
👤
Senior🎓
Top SchoolAmanda Dsouza is a Staff Applied Research Scientist with 11 years of experience building and shipping NLP and ML solutions across startups and enterprise teams, currently at Snorkel AI after roles at Jasper and Fractal Analytics. She blends applied research with production-first engineering—leading teams on text mining and large language model projects and contributing to core open-source tooling like scikit-learn and data-prep libraries. A Georgia Tech MS in Computer Science (ML) informs her work in RL, NLP, and robust ML engineering, where she focuses on reliability, missing-value handling, and model evaluation. Based in Seattle, she pairs rigorous academic training with a practical track record of improving library quality and reproducibility in widely used ML projects.
11 years of coding experience
16 years of employment as a software developer
Master's degree, Computer Science, Master's degree, Computer Science at Georgia Institute of Technology
Github Skills (14)
scikit-learn10
data-preprocessing10
scikit10
pandas10
machine-learning10
dataprep10
data-cleaning10
python10
data-science10
data-analysis10
pytest9
numpy9
data-wrangling8
documentation7
Programming languages (2)
Github contributions (5)
skrub-data/skrub
Jun 2020 - Nov 2020
Prepping tables for machine learning
Role in this project:
Data Scientist
Data ScientistContributions:10 reviews, 24 commits, 6 PRs in 5 months
Contributions summary:Amanda primarily contributed to the `dirty_cat` library, focusing on the `SimilarityEncoder` and `TargetEncoder` classes. Their commits demonstrate work on handling missing values, including implementing error handling and imputing missing data. The user added tests for missing value scenarios and addressed parameter issues within the encoders, indicating a focus on code quality and robustness.
autoencoderdata-preprocessingdatadata-sciencetabular-data
scikit-learn/scikit-learn
Jun 2020 - Nov 2021
scikit-learn: machine learning in Python
Role in this project:Data Scientist
Data ScientistContributions:8 reviews, 12 commits, 15 PRs in 1 year 5 months
Contributions summary:Amanda contributed to the standardization of documentation formatting within the scikit-learn library, specifically addressing default values in various machine learning modules. They fixed issues related to incorrect cluster generation in AffinityPropagation when using float32 input and addressed an overflow issue in IncrementalPCA on Windows. Their work also included adding support for fit_params in learning_curve, permutation_test_score, and validation_curve, enhancing the flexibility of model evaluation. They also added support for categories with missing values to fetch_openml when using as_frame=True.
data-analysispythonstatisticsdata-sciencelearn-machine-learning
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.