Built with Love and Scale by ex-Netflix, Databricks, and Tensorflow Engineers

Videos

• Continuously Train & Deploy Spark ML and Tensorflow AI Models from Jupyter Notebook to Production (StartupML Conference Jan 2017)
• Recent Advancements in Data Science Workflows: From Jupyter-based Notebook to NetflixOSS-based Production (Big Data Spain Nov 2016)

Reference Architecture

100% Open Source

Completely Kubernetes and Docker Based

PMML + Native C++/Java Code Generation for Optimized CPU/GPU Performance

Deploy directly from Jupyter Notebook to Hybrid-Cloud Microservices

NetflixOSS-based Microservices for High Scalability and Availability

Continuous Train/Deploy Spark ML and Tensorflow AI Pipelines to Production

KINESIS

Environment Setup

Pipeline Mini (Local Laptop)

• Small Kubernetes cluster for your laptop

Pipeline AWS

• Kubernetes cluster on any cloud provider or on-premise

Pipeline Google

• Kubernetes cluster across many cloud providers and/or on-premise

24x7 Global Support for Everyone!

• Create a Ticket
• Email [email protected]

Additional Resources

Training Repo and Workshop Schedule

Global Advanced Spark and Tensorflow Meetup

Slideshare Slides

YouTube Videos

Docker Images