Machine Learning Infrastructure, Software Engineer (Principal/Staff/Senior)
San Francisco, CA
At Instacart, we use machine learning and internet scale data to elevate customer experience, improve efficiency and reduce cost in e-commerce, advertising, and fulfillment. For example, we build large, distributed machine learning models for personalization and recommendation. We're looking for experienced system and infrastructure engineers with background in building ML platforms to join our fast-moving team.
ABOUT THE JOB
There is tremendous opportunity in front of us, and joining now gives you a chance to grow your career and interests as we succeed.
- You will design solutions for machine learning infrastructure that will support the current and future needs of our business.
- You will apply your system software engineering skills and machine learning knowledge to build scalable and easy-to-use machine learning workflows to improve the scalability and efficiency in launching machine learning solutions.
- You will enable machine learning teams to perform scalable training, evaluation, and inference in the cloud and in client-side infrastructure.
- You will enable software engineers across the company to use machine learning solutions in their work.
- You will work closely with related teams, including Search, Ads, Personalization & Recommendation, Catalog, etc. You will have tremendous ownership and responsibility for managing things directly.
- Background in Computer Science, Math, Statistics, or a related field.
- 5+ years of industry experience building ML infrastructure at scale.
- Proficient in Python or C++. Experience writing and maintaining high-quality production code.
- Knowledge of machine learning (particularly hyperparameter tuning and debugging).
- Experience serving models using a variety of ML model frameworks like Tensorflow, PyTorch, Sci-kit Learn, etc.
- Experience using AWS to build data-intensive infrastructure.
- Experience building platforms around publicly available ML platforms like Sagemaker, Kubeflow, MLflow, Horovod, etc.