Software Engineer (Ray Data) - Anyscale
About Anyscale:
At Anyscale, we're on a mission to democratize distributed computing and make it accessible to software developers of all skill levels. We’re commercializing Ray, a popular open-source project that's creating an ecosystem of libraries for scalable machine learning. Companies like OpenAI, Uber, Spotify, Instacart, Cruise, and many more, have Ray in their tech stacks to accelerate the progress of AI applications out into the real world.
With Anyscale, we’re building the best place to run Ray, so that any developer or data scientist can scale an ML application from their laptop to the cluster without needing to be a distributed systems expert.
Proud to be backed by Andreessen Horowitz, NEA, and Addition with $250+ million raised to date.
Anyscale is based in San Francisco, CA. Employees are required to come in office 3x a week.
About the role:
Ray aims to provide a universal API for building distributed applications (e.g. a machine learning pipeline of feature engineering, model training, and evaluation). Data is usually a core element connecting these different stages, and therefore plays a critical role in Ray’s usability, performance, and stability. We are looking for strong engineers to build, optimize, and scale Ray’s Datasets library and data processing capabilities in general.
About the Ray Data team:
The Ray Data team currently develops and maintains the Ray Datasets library, which is already powering critical production use cases (e.g. large scale data compaction at Amazon, and ML pipeline at Alibaba). Ray Datasets is a Python library built on top of Apache Arrow and Ray Core (Ray’s C++ backend), and the Ray Data team interacts closely with Ray Core components including the scheduler and the memory & I/O subsystems. The Ray Data team also works closely with Ray’s ML libraries including Train, RLlib, and Serve.
A snapshot of projects you will work on:
- Performance of Ray Datasets at large scale (leveraging Arrow primitives, optimizing Ray object manager, etc.)
- Integration with ML training and data sources
- Stability and stress testing infrastructure
- Lead future work integrating streaming workloads into Ray such as Beam on Ray
- Differentiate Data operations in Anyscale hosted Ray service
Last updated: 11 hours ago
© 2021 - 2022