Matt Gardner
Research Scientist at Scaled Cognition
Irvine, California, United States
Summary
🤩
Rockstar🎓
Top SchoolMatt Gardner is a research scientist with 13 years of experience bridging academic NLP research and production-grade ML systems, currently at Scaled Cognition after roles at Microsoft and the Allen Institute for AI. He holds a PhD in Language Technologies from CMU and has driven research and engineering work on conversational AI and language models. Matt contributes to influential open-source projects like AllenNLP and Keras, improving data handling, model robustness, and documentation for the broader NLP community. His background includes principal research responsibilities at Microsoft (Semantic Machines) and hands-on fixes across model APIs and serialization that signal a blend of deep theory and pragmatic engineering. Based in Irvine, California, he pairs rigorous research instincts with attention to code clarity and maintainability—often surfacing subtle bugs and improving developer usability in widely used libraries. Colleagues describe him as someone who comfortably moves between publishing research and shipping reliable tooling that makes that research usable.
13 years of coding experience
8 years of employment as a software developer
Doctor of Philosophy (Ph.D.), Language technologies, Doctor of Philosophy (Ph.D.), Language technologies at Carnegie Mellon University
English, Spanish, slovene
Github Skills (18)
pytorch10
python10
data-science10
machine-learning10
keras10
deep-learning10
tensorflow10
neural-network10
nlp10
model-driven9
testing9
modeling9
model-driven-development9
model-building9
data-structures8
Programming languages (13)
Github contributions (5)
allenai/allennlp
Jun 2017 - Nov 2020
An open-source NLP research library, built on PyTorch.
Role in this project:
Data Scientist
Data ScientistContributions:1 release, 39 reviews, 273 commits in 3 years 5 months
Contributions summary:Matt contributed to documentation and performed namespace fixes for field classes within the `allennlp` research library, built on PyTorch. Their work included code modifications, especially within `data/fields`, such as `label_field.py`, `tag_field.py`, and `index_field.py`, indicating a focus on enhancing data handling and structure for NLP tasks. Additionally, the user improved docstrings for methods in `vocabulary.py`, showcasing attention to code clarity and usability within the AllenNLP project.
nlppytorchtransformerspythonbert
allenai/allennlp-models
Mar 2020 - Oct 2020
Officially supported AllenNLP models
Role in this project:ML Engineer
ML EngineerContributions:5 reviews, 6 commits, 16 PRs in 7 months
Contributions summary:Matt contributed to the AllenNLP models repository by fixing bugs and improving existing functionalities. They addressed issues related to ELMo embedder API changes, coreference interpretation tests, and sentiment analysis. Additionally, the user added a new test for sentiment analysis and utilized standard utility functions in a masked language model, demonstrating a focus on model accuracy and maintainability. They also corrected a capitalization typo in a configuration file.
nlppytorchldabertword-embeddings
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.