Cohere For AI (C4AI) is the dedicated research arm of Cohere. We work at the frontier of AI progress with the goal of solving cutting-edge scientific problems. We see contributions to top-tier conferences and publications in journals as an important part of our work, but also drive the creation of new research spaces and breakthroughs that change where, how and by whom research is done. Our mission is to solve complex machine-learning problems at the edge of what is currently technically possible. Cohere For AI has both a full-time research staff which builds the next generation of large scale machine learning models and an open science initiative that supports collaborative research across institutions and creates paths of access for independent researchers.
In this role, as a Machine learning Engineer, you work mainly on creating great libraries and usability for open research releases. This includes trained models, large-scale datasets, tools/libraries, and accessible walk-throughs of scientific research and technical breakthroughs. You are excited to be creative and create artifacts that make research models and open weights accessible to users and contributors of the broad open-source machine-learning ecosystem.
An essential aspect of this role is engaging with the wider open-source ML ecosystem, interacting with and learning from its users and contributors. Your responsibilities will include collaborating with researchers, ML practitioners and data scientists, making our research models accessible and intuitive to use on Huggingface, Kaggle answering queries and encouraging contributions and research extensions to released work via GitHub and our open science community.
As a machine learning engineer focused on open source, you will:
Support open release of scientific artifacts, make models highly optimized and usable for developer hardware.
Establish best practices and processes for open source releases. You are excited to make our releases accessible and easy to use to the widest possible range of users by testing regularly usability, creating easy to use guides and promoting best practices in responsible use.
Review and triage public issues, questions, and pull requests.
Develop and integrate software to support the open source release process.
Show creativity with how you make our models and research insights accessible and delightful to a wide variety of developers.
You may be a good fit if you:
3 years of model training, deployment, and maintenance in a production environment.
Strong skills in NLP and deep learning.
Experience scaling products at hyper-growth startup.
Strong written and verbal communication skills.
Proficiency in Python and related ML frameworks such as Tensorflow, TF-Serving, JAX, Pytorch and XLA/MLIR.
Excitement and interest in efficiency techniques to make open science more usable under compute constraints.
Experience using large-scale distributed training and inference.
Strong mentorship, communication, and problem-solving skills.
A demonstrated passion for applied ML models and products.