Joanna Lowry-duda
Software Engineer, Machine Learning at Facebook
Boston, Massachusetts, United States
Summary
🤩
Rockstar🎓
Top SchoolJoanna Lowry-duda is a software engineer specializing in machine learning with 11 years of experience, currently building ML systems at Facebook in Boston. Her background blends computational linguistics (M.A., Brandeis) and Slavic languages (B.A., Jagiellonian), giving her a rare dual expertise in language theory and production NLP. Before Facebook she led research and engineering efforts at Luminoso, and has contributed significant back-end evaluation work to the well-known ConceptNet project—adding SemEval2017 support and robust word-similarity tests. She focuses on turning linguistic insight into scalable evaluation and inference pipelines, with strengths in refactoring legacy code and designing reproducible evaluation workflows. Colleagues describe her as a thoughtful engineer who bridges research rigor and pragmatic product delivery.
11 years of coding experience
5 years of employment as a software developer
Bachelor of Arts (B.A.), Slavic Languages, Literatures, and Linguistics, 5.0, Bachelor of Arts (B.A.), Slavic Languages, Literatures, and Linguistics, 5.0 at Jagiellonian UniversityMaster of Arts (M.A.), Computational Linguistics, Master of Arts (M.A.), Computational Linguistics at Brandeis University
English, Croatian, Polish
Github Skills (5)
pandas10
python10
numpy9
scipy9
machine-learning8
Programming languages (4)
Github contributions (5)
commonsense/conceptnet5
Jan 2017 - Sep 2020
Code for building ConceptNet from raw data.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:204 commits, 83 PRs, 112 pushes in 3 years 8 months
Contributions summary:Joanna significantly contributed to the evaluation component of the ConceptNet5 project, specifically by adding and refining various word similarity evaluation methods. They introduced support for SemEval2017 evaluations, both monolingual and cross-lingual, by modifying the existing `wordsim.py` file. The commits also involved code refactoring, including separating functionalities and streamlining the overall evaluation process. These changes included adding tests for evaluating word correlations and simplifying the structure of how scores are obtained, enhancing the functionality and analytical capabilities of the project.
pytorchraw-datadeep-learningsemanticsnatural-language-processing
LuminosoInsight/exquisite-corpus
May 2018 - Sep 2020
Put together a multilingual corpus from a variety of sources. Used for wordfreq and word embeddings.
Contributions:1 review, 51 commits, 24 PRs in 2 years 4 months
nlpword-embeddingsembeddingscorpuscorpus-linguistics
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.