Google Cloud Next – The Best Parts!
By Karthik Muthupalaniappan
Google certainly knows how to put on a great conference – like Google Cloud Next, which I attended in March. If you’re looking for a showcase of Google Cloud’s product offerings, this is the place. With so much to learn and digest in three days, this blog post covers the key things I took from the conference.
Along with product features, Google Cloud Next shows how these offerings can help build and deliver modern software applications. The topics were wide-ranging: Machine Learning, Microservices, DevOps, Conversational APIs, Developer Productivity were some of the key content areas.
If there was a one single widely discussed topic in this conference, it was Machine Learning – the next big thing in the world of software engineering. Machine Learning can be crudely defined as the scientific approach to teaching computers to be smarter. I enjoyed delving into the basics of Machine Learning, Deep Neural Networks and Machine Learning APIs that are part of Google Cloud.
One simple use case of Machine Learning is the smart reply feature of Google Inbox, an excellent feature I use personally. The capabilities of Tensorflow, Google’s flagship Machine Learning framework, were also showcased.
Another interesting talk focused on how Google did on-device machine learning for Android. The challenge of running and processing Machine Learning models on a mobile or embedded device is infinitely greater than doing it in the cloud. There was a glimpse into how on-device Machine Learning is possible using Tensorflow Mobile.
Google provides a suite of powerful Machine Learning APIs like the Cloud Machine Learning API, Natural Language and Translation API, Translation API, Vision and Video Intelligence API. You can learn more and play around with all of these APIs on Google’s Cloud Machine Learning Services page.
The Natural Language and Translation API is a good choice for someone looking to wrap their head around how Machine Intelligence is applied to processing Natural Language text or queries. The explorer for this API is available on the Google Developer’s site.
These are a couple of talks I would recommend watching:
Microservices and Kubernetes
Kubernetes is Google’s container orchestration and management platform. There were a number of talks centered around how Microservices can be built and deployed at scale using Kubernetes. Container-based Microservice deployments are becoming popular given the speed and elasticity the approach provides. The Google Cloud container builder is a service that allows you to build containers from application code sitting in Google Cloud and persist these containers in the Google Container Registry, the equivalent of DockerHub.
Another key building block for Microservices in Google Cloud is gRPC, which is an open source RPC framework Google has built and maintained over the years to facilitate seamless inter-service communication. gRPC is built on top of Protocol Buffers, a serialization framework written by Google again.
Interested in learning more? Watch Building high performance microservices with Kubernetes, Go, and gRPC.
What are conversational APIs? These are interfaces that are behind the development of Bot applications such Chat Bots, Messenger Bots, etc. To give you a simple example, a bot that can converse with you in your favorite messenger apps such as Facebook Messenger or Skype and help you with something like checking your bank account balance or paying your electricity bill, etc. These applications have numerous use cases in the real world and are gaining a lot of popularity.
Api.ai is basically conversational interfaces as a service. This really opens up multiple possibilities for interacting with systems such as your bank, financial applications, government agencies, etc. Watch this demo of building a home automation bot in Skype using Api.ai: Your next app might be a bot! Building conversational UX with API.AI.
Enjoy, and I hope you find these talks as informative as I did.