Notebooks has become a ubiquitous IDE for data scientists across all industries. But the most ML platform APIs are not friendly nor support notebook based development natively. This talk shows how a data scientist can use Kubeflow Fairing SDK to seamlessly build, train, and deploy their machine learning models to their on-prem cluster and with just one switch of flag to the managed google cloud environment.
The talk covers machine learning best practices in hybrid cloud to minimize development time by training locally with sampled data and to minimize infrastructure cost by using GPUs/TPUs only during the course of model training.
Finally the talk will cover how to take your trained model into production by creating an online prediction endpoint that can do custom python preprocessing. All these tasks are done from a notebook without leaving the notebook interface and also not using any CLIs and using very data scientist friendly interface.
Karthik is a machine learning engineer leading the data science experience for Kubeflow and working primarily on Kubeflow Fairing project. Before Google, Karthik was part of ML platform team at Uber self driving team and also led a team of data scientists in Uber's Risk team. He is passionate about making data scientists productive.
5:45 - 6:30 PM Networking/Social/Food
6:30 - 6:35 PM Introduction
6:35 - 7:30 PM Talk and Q&A
7:30 - 8:00 PM Networking/Social/Closing