Building the modern, human-centric ML infrastructure stack.

Our roots are at Netflix, where we started Metaflow - an open-source framework that helps data scientists and ML engineers develop and deliver real-life ML projects. At Netflix, we saw that successful data science projects were delivered by data scientists and ML engineers who can work on end-to-end workflows independently, focusing more on data science, less on engineering.

ML applications come in all shapes and sizes. Sophisticated companies experiment with new ideas constantly and deploy them to production without friction. These applications are developed by a diverse group of domain experts using a rapidly evolving set of libraries. There isn't a single cookie-cutter way of delivering business value through ML.

There is an increasing need for infrastructure that provides a solid foundation for ML applications to both - experiment with new ideas and deploy promising prototypes quickly. This infrastructure should allow data scientists to execute millions of compute-intensive jobs in the cloud, handle large amounts of data, orchestrate and operate complex applications, and iterate on projects quickly and confidently. This infrastructure should integrate seamlessly into existing business systems and processes since ML can’t be an island. Most importantly, this infrastructure should make its users exceptionally productive and happy, regardless of their technical background.

We are committed to building this infrastructure in open-source - in collaboration with a quickly growing community of sophisticated companies, infrastructure teams, data scientists, and ML engineers. We invite you to join our community to learn more, get support for Metaflow, and participate in using and building modern, human-centric ML infrastructure. Also, if this vision resonates with you, we are hiring.

Outerbounds ©2021