Pierre Andrews
Research Engineer at Meta
Mountain View, California, France
Summary
🤩
Rockstar🎓
Top SchoolPierre Andrews is a research engineer with 18 years of experience building scalable ML infrastructure and data systems, currently working at Meta in Mountain View. His career blends deep research (PhD in computer science) with hands-on engineering across Scala/Akka/Kafka back ends, Hadoop workflows, and interactive front-ends, having led high-concurrency social data platforms at Quantifind and machine learning infra at Facebook. He contributes to open-source AI tooling—improving preprocessing modularity in Facebook Research’s widely used fairseq—demonstrating a knack for optimizing large-data pipelines for parallelism. Pierre’s work also spans civic tech and NLP research, from parliamentary data transparency projects to natural language interfaces, reflecting a long-standing interest in making complex data accessible. Fluent across academic and production environments, he combines rigorous research instincts with pragmatic system design.
18 years of coding experience
9 years of employment as a software developer
Ph.D., Ph.D. at University of Yorkbaccalaureate, baccalaureate at College du LemanBsc+Msc, Bsc+Msc at EPFL (École polytechnique fédérale de Lausanne)
English, French, Italian
Github Skills (11)
pytorch10
python10
fairseq10
file-access9
fileio9
file-processing9
python-multiprocessing9
file-handling9
multi-process9
multiprocessing9
artificial-intelligence8
Programming languages (11)
Github contributions (5)
facebookresearch/fairseq
Feb 2021 - Feb 2022
Facebook AI Research Sequence-to-Sequence Toolkit written in Python.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:2 reviews, 14 commits, 3 PRs in 11 months
Contributions summary:Pierre contributed to the `fairseq` repository, a sequence-to-sequence toolkit for AI research. Their primary contribution was extracting file chunking logic from the preprocessing script, improving code modularity. The code changes involved modifying the `preprocess.py` and creating a new utility file `file_chunker_utils.py`. This likely optimized the data preprocessing pipeline for tasks like machine translation by enabling parallel processing of large datasets.
pytorchnlpsequencepythontransformer-architecture
quantifind/KafkaOffsetMonitor
Jan 2015 - Jul 2020
A little app to monitor the progress of kafka consumers and their lag wrt the queue.
Contributions:1 release, 12 PRs, 14 pushes in 5 years 6 months
monitorkafka-consumersconsumerswrtkafka
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.