Max Woolf
Senior Data Scientist
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolMax Woolf is a Senior Data Scientist based in San Francisco with 13 years of experience turning messy data into clear, actionable stories. After starting in software QA at Apple, he transitioned to data roles at BuzzFeed where he progressed from contract Data Scientist to senior lead, blending product-minded analysis with production ML. He maintains the popular minimaxir open-source projects—ranging from GPT-2 tooling (gpt-2-simple, aitextgen, textgenrnn) to useful utilities like the Big List of Naughty Strings—which reflect a practical focus on reproducible model training and robust data pipelines. Max combines strong visualization chops (ggplot2 and interactive plotting) with hands-on ML engineering, feature engineering, and automation work, and he often publishes notebooks that make social media and engagement data easy to explore. Notably, his tooling emphasizes usability—features like nucleus sampling, Drive integration, and automated model/code generation show a pragmatic drive to lower barriers for others to train and deploy text models.
13 years of coding experience
12 years of employment as a software developer
Bachelor of Science - BS Computing and Information Technology, Bachelor of Science - BS Computing and Information Technology at Carnegie Mellon University
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (46)
transformers10
text-generation10
data-visualizations10
json10
tokenize10
pytorch10
xgboost10
python10
image-processing10
r10
scikit10
pandas10
machine-learning10
rnn-model10
data-visualisation10
Programming languages (9)
Github contributions (5)
minimaxir/automl-gs
Nov 2018 - Apr 2019
Provide an input CSV and a target field to predict, generate a model + code to run it.
Role in this project:
Data Scientist
Data ScientistContributions:1 release, 94 commits, 6 PRs in 4 months
Contributions summary:Max contributed significantly to the development of an automated machine learning system, focusing on model generation and data processing. Key changes included the implementation of numeric feature strategies such as quantiles and percentiles, reflecting an interest in feature engineering techniques. They also worked on adapting the pipeline to different frameworks, including the addition of text processing features using CountVectorizer, showing a focus on flexibility and model versatility. They were responsible for incorporating a more efficient approach to hyperparameter search, improving the trial selection speed.
csvpythontidymodelspredictattribution
minimaxir/textgenrnn
Jun 2017 - Jul 2020
Easily train your own text-generating neural network of any size and complexity on any text dataset with a few lines of code.
Role in this project:ML Engineer
ML EngineerContributions:12 releases, 136 commits, 26 PRs in 3 years 1 month
Contributions summary:Max implemented a text-generating neural network using Keras and TensorFlow. Their contributions focused on defining the model architecture, incorporating embedding layers, and integrating recurrent neural networks (LSTMs). The user also worked on training the model, including data encoding and the implementation of a learning rate scheduler, as well as adding functionalities to generate text.
pytorchnlppythoncomplexitylines
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.