Martin Bomio
Senior Staff Machine Learning Engineer at Spotify
United States
Summary
🤩
Rockstar🎓
Top SchoolMartin Bomio is a Senior Staff Machine Learning Engineer with a decade of experience building scalable ML and data infrastructure at Spotify, rising through roles from Data Engineer to senior technical leadership. He specializes in production ML pipelines and data ingestion performance—contributing to high-profile open-source projects like TFX and Ray to add Parquet support and accelerate TFRecord reading via tfx-bsl. Comfortable across backend, data engineering, and ML model serving, he has a track record of practical optimizations (e.g., skewed join fixes and TensorFlow sequence example support) that improve throughput at scale. He also brings teaching experience from Universidad de Montevideo, reflecting an ability to communicate complex big-data concepts to engineers and students alike.
10 years of coding experience
12 years of employment as a software developer
Engineer's Degree Telematics Engineering, Engineer's Degree Telematics Engineering at Universidad de MontevideoSecundaria, Secundaria at St. Patrick's CollageBachillerato, Bachillerato at Preuniversitario CarrascoPrimaria, Primaria at St. Patrick's College
Spanish, English
Stackoverflow
11reputation
2kreached
0answers
4questions
Github Skills (36)
bigquery10
python10
data-science10
data-set10
data-engineering10
machine-learning10
google-cloud-dataflow10
filetype10
file-conversion10
file-operations10
data-model10
file-format10
scala10
data-preprocessing10
tensorflow10
Programming languages (6)
Github contributions (5)
spotify/scio
Nov 2017 - Nov 2020
A Scala API for Apache Beam and Google Cloud Dataflow.
Role in this project:
Back-end Developer & Data Engineer
Back-end Developer & Data EngineerContributions:1 review, 8 commits, 10 PRs in 3 years
Contributions summary:Martin contributed to the `scio` project, a Scala API for Apache Beam and Google Cloud Dataflow, focusing on enhancements to TensorFlow integration. Their work includes adding support for sequence examples within the TensorFlow ecosystem, reworking TensorFlow predict operations, and refactoring existing TensorFlow save functions. Furthermore, the user addressed issues related to skewed joins, demonstrating expertise in data processing optimizations.
apidatabeambatchbigquery
ray-project/ray
Mar 2023 - Mar 2025
Ray is an AI compute engine. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.
Role in this project:Data Scientist
Data ScientistContributions:38 reviews, 15 PRs, 55 comments in 2 years
Contributions summary:Martin significantly contributed to improving the performance of reading TFRecord datasets within the Ray framework, focusing on a performant approach using the tfx-bsl decoder. They implemented changes in the `TFRecordDatasource` class, incorporating TFXReadOptions to optimize data reading and schema inference. These modifications enable faster processing of large datasets by utilizing the tfx-bsl decoder and supporting automatic schema inference when a tf_schema is not provided. Further contributions involved enhancing preprocessors, specifically to address and fix several issues within the concatenator and imputer preprocessors.
pythonconsistsruntimetensorflowserving
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.