Ian Simon
Music & ML Nerd at Google
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolIan Simon is a Music & ML engineer with a Ph.D. in Computer Vision and 13 years of experience applying machine learning to music, audio, and image problems. He helped create MySong (later Microsoft Songsmith), built music recognition for Bing Audio, and led Smule’s music recommendation engine before joining Google’s Magenta project to advance ML-driven music and art creation. His open-source contributions span influential projects—from GPUImage filters for complex iOS image processing to core Magenta work adding chord, lead-sheet, and transcription capabilities—showing fluency across research, backend systems, and production code. He combines academic rigor in large-scale visual information extraction with practical expertise in musical structure and transformer-based transcription models, often surfacing subtle data-handling fixes that improve real-world model robustness.
13 years of coding experience
Ph.D, Computer Science, Ph.D, Computer Science at University of WashingtonB.S., Computer Science, B.S., Computer Science at Washington University in St. Louis
Stackoverflow
43reputation
220reached
1answer
2questions
Github Skills (31)
music-generation10
javascript10
ios10
python10
image-processing10
tensorflowjs10
music-theory10
machine-learning10
gpuimage10
data-preprocessing10
transformer-models10
google-colab10
tensorflow10
google-colaboratory10
nlp10
Programming languages (7)
Github contributions (5)
magenta/mt3
Oct 2021 - Jan 2023
MT3: Multi-Task Multitrack Music Transcription
Role in this project:
ML Engineer
ML EngineerContributions:39 commits, 1 PR, 3 pushes in 1 year 3 months
Contributions summary:Ian primarily contributes to the MT3 project, focusing on music transcription using transformer models. Their work involves modifications to model configurations, including adjustments to loss functions and related parameters. They add dependencies and configure the environment for the project, alongside adding and improving Colab notebooks for demo and testing purposes.
multi-taskmusic-transcriptionmiditranscriptionaudio
magenta/magenta
Sep 2016 - Aug 2022
Magenta: Music and Art Generation with Machine Intelligence
Role in this project:Back-end Developer
Back-end DeveloperContributions:6 releases, 1 review, 184 commits in 5 years 11 months
Contributions summary:Ian's contributions centered on enhancing the codebase for music and art generation with machine intelligence, focusing on backend logic. They added essential support for chord progression implementation, lead sheets, and handling data time signature anomalies, demonstrating an understanding of musical structure. The user's commits also included refactoring of the underlying event sequence management, consolidating functionalities across melodies, chord progressions, and lead sheets. They were instrumental in implementing core capabilities related to data handling and music modeling.
artmusic-generationgenerative-artmagentamachine-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.