Josh Devins
Senior Principal Engineer, Search And Machine Learning at Elastic
Berlin, Germany
Summary
🤩
Rockstar🎓
Top SchoolJosh Devins is a Senior Principal Engineer specializing in search and machine learning, bringing two decades of industry experience and 14 years in senior engineering roles across North America and Europe. Based in Berlin and leading ML for Search at Elastic, he focuses on neural information retrieval, NLP, LLMs and RAG systems, blending hands-on research with product-focused engineering. He builds and mentors cross-functional teams to deliver distributed data products and high-quality software, often working at the intersection of data science and engineering. An active open-source contributor, he has implemented PyTorch transformer support and practical ML tooling for Elasticsearch (notably in elastic/eland and search-focused examples), demonstrating deep expertise in productionizing retrieval models. His academic background in computer science with a philosophy minor underpins a pragmatic, systems-oriented approach to complex, data-driven problems.
13 years of coding experience
Bachelor of Applied Science - BASc, Major: Computer Science, Minor: Philosophy, Bachelor of Applied Science - BASc, Major: Computer Science, Minor: Philosophy at Simon Fraser University
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (21)
transformers10
pytorch10
python10
data-science10
amazon-elasticsearch10
query-optimization10
machine-learning10
elasticsearchquery10
aws-elasticsearch10
elasticsearch10
jupyter-notebook10
nlp10
elasticsearch-api10
pandas7
data-analysis6
Programming languages (9)
Github contributions (5)
elastic/eland
Nov 2019 - Dec 2021
Python Client and Toolkit for DataFrames, Big Data, Machine Learning and ETL in Elasticsearch
Role in this project:
ML Engineer
ML EngineerContributions:38 reviews, 8 commits, 13 PRs in 2 years
Contributions summary:Josh primarily contributed to the integration of PyTorch machine learning models within the Elasticsearch ecosystem. They implemented initial support for PyTorch models, specifically focusing on transformers, including wrappers for sentence-transformers and DPR models. Their work involved defining traceable models, adding padding functionality for tokenization, and upgrading the PyTorch dependencies to the latest versions to address compatibility issues and leverage recent features.
data-analysisdataframespythondata-sciencedataframe
elastic/examples
Jun 2020 - Feb 2021
Home for Elasticsearch examples available to everyone. It's a great way to get started.
Role in this project:Data Scientist
Data ScientistContributions:22 commits, 6 PRs, 43 pushes in 8 months
Contributions summary:Josh's commits center around implementing and optimizing machine learning models for search relevance within the context of the Elasticsearch platform. Their work involves using the `eland` library and Jupyter notebooks to analyze search metrics and fine-tune query parameters. They are focused on document ranking tasks, and experimenting with document expansion techniques, demonstrating a strong understanding of data-driven query optimization.
elasticgreatget-startedelasticsearch
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.