Google Cloud Platform Podcast

Amy Unruh and Sara Robinson join the podcast this week to talk with Mark and Melanie about the alpha launch of Cloud AutoML Vision.

Cloud AutoML is a suite of products enabling developers with limited ML expertise to build high quality models using transfer learning and Neural Architecture Search techniques. AutoML Vision is the first product out the gate with a focus on making it easy to train customized vision models.

Amy is a developer relations engineer for the Google Cloud Platform, where she focuses on machine learning and data analytics as well as other Cloud Platform technologies. Amy has an academic background in CS/AI and has also worked at several startups, done industrial R&D, and published a book on App Engine.

Sara is a developer relations engineer on Google's Cloud Platform team, focusing on big data and machine learning. She worked on providing initial product feedback and building a demo for the AutoML Vision launch.

• Google Brain Looking Back on 2017 blog
  • Shout-out to Kaz Sato for his TensorFlow Rock Paper Scissors example
• Running dedicated game servers in Kubernetes Engine blog
• Kaggle Learn site

Honorable mention… - Scientists put a worm brain in a lego robot blog

• Cloud AutoML: Making AI accessible to every business blog
• Cloud AutoML Vision site
• Cloud AutoML Vision Access Request | Whitelist Application form
• Cloud images example video
  • Shout-out thanks to Rob Carver for domain expertise in helping label cloud images.
• Coastline images example readme and filenames csv
• Using Machine Learning to Explore Neural Network Architecture blog
• Learning Transferable Architecture for Scalable Image Recognition arXiv paper
• Neural Architecture Search with Reinforcement Learning arXiv paper
• Progressive Neural Architecture Search arXiv paper
• Learning2learn video
• Cloud Vision site docs

How does someone in academia get GCP credits?

• Google Cloud Platform Education Grants site

Melanie is speaking at AI Congress in London Jan 30th and she will be at FOSDEM in Brussels in Feb.

Mark will be at the Game Developer's Conference | GDC in March.

Bringing you a special second episode this week with Matt Linton and Paul Turner sharing insights with Mark and Melanie about the CPU vulnerabilities, Spectre & Meltdown, and how Google coordinated and managed security with the broader community. We talked about how there has been minimal to no performance impact for GCP users and GCP’s Live Migration helped deploy patches and mitigations without requiring maintenance downtime.

Due to the special nature, no cool things or question included on this podcast.

Amy Unruh and Sara Robinson join the podcast this week to talk with Mark and Melanie about the alpha launch of Cloud AutoML Vision.

Cloud AutoML is a suite of products enabling developers with limited ML expertise to build high quality models using transfer learning and Neural Architecture Search techniques. AutoML Vision is the first product out the gate with a focus on making it easy to train customized vision models.

Launchpad Studio, a product development acceleration program focused on helping machine learning startups iterate quickly, fail fast, and collaborate on best practices.

Malika Cantor and Peter Norvig talk with Mark and Melanie this week about how the Launchpad Studio program is helping startups overcome data, expertise and tooling barriers by providing access to talent and resources and building universal best practices.

Malika is the Global Lead for Google Launchpad Studio. Launchpad is the acceleration engine of Google - running a number of accelerator programs focused on supporting the global startup ecosystem. Prior to joining Google, Malika was a co-founder and partner at Comet Labs, a venture capital firm and experimental research lab focused on investing and supporting applied AI startups. She has worked with founders for around 6 years, in London, Beijing, Singapore, Toronto, and Silicon Valley.

Peter Norvig is a Director of Research at Google; previously he directed Google's core search algorithms group. He is a fellow and councilor of the Association for the Advancement of Artificial Intelligence and co-author, with Stuart Russell, of Artificial Intelligence: A Modern Approach, now the leading college text in the field. He was head of the Computational Sciences Division (now the Intelligent Systems Division) at NASA Ames Research Center.

• Introducing Preemptible GPUS: 50% off blog
• How We Implemented a Fully Serverless Recommender System Using GCP blog
• Awesome lists repo:
  • Awesome Google Cloud Platform repo
  • Awesome Kubernetes repo
  • Awesome TensorFlow repo
  • Awesome Firebase repo

• Launchpad Studio site
• Verily site
• DeepMind site
• Cloud AI site
  • Cloud Machine Leanring Engine site
• TensorFlow site
• Android Things site

How does a startup get GCP credits?

• Google Cloud Platform Startup Program site

• What Google Cloud, G Suite and Chrome customers need to know about the industry-wide CPU vulnerability blog
• Google Security Blog, Today's CPU vulnerability: what you need to know blog
• ProjectZero News and Updates blog

San Francisco

Launchpad Studio, a product development acceleration program focused on helping machine learning startups iterate quickly, fail fast, and collaborate on best practices.

Malika Cantor and Peter Norvig talk with Mark and Melanie this week about how the Launchpad Studio program is helping startups overcome data, expertise and tooling barriers by providing access to talent and resources and building universal best practices.

This week we get the band back together! Francesc Campoy Flores rejoins the show along with Director of Google Cloud Developer Relations Greg Wilson to talk all about 2017 and Google Cloud with Mark and Melanie

Francesc Campoy Flores is the VP of Developer Relations at source{d}, He's also a Gopher, Catalan, LGBTQIA advocate, previous Google employee (and Podcast host), and creator of the Just For Func YouTube series!

Greg Wilson is the Director of Google Cloud Developer Relations, overseeing developer relations work across both G Suite and Google Cloud Platform.

• Jeff Dean's talk at NIPS on ML for Systems and Systems for ML sides
• The Case for Learned Index Structures paper
• KubeFlow github hackernews
• Manage Google Kubernetes Engine from Cloud Console dashboard, now generally available blog

• Top 5 Downloaded Episode of 2017
  • #88 Kubernetes 1.7 with Tim Hockin
  • #91 The Future of Media with Machine Learning with Amit Pande
  • #93 What's AI with Melanie Warrick
  • #75 Container Engine with Chen Goldberg
  • #100 Vint Cerf: past, present, and future of the internet
• Greg's Favourites
  • #57 Pokémon GO with Edward Wu, Director of Software Engineering at Niantic
  • #68 The Home Depot with William Bonnell
  • #86 Broad Institute and Platinum Customers with Lukas Karlsson and Mike Altarace
• Francesc's Favourites
  • #62 Cloud Spanner with Deepti Srivastava
• Mark's Favourites
  • The SRE Category on GCP Podcast
• Melanie's Favourites
  • #57 Pokémon GO with Edward Wu, Director of Software Engineering at Niantic
• Favourite announcements, products and more at Google Cloud Platform
  • Cloud Spanner
  • Cloud Machine Learning Engine
  • TensorFlow
  • GCE Virtual Machines, e.g. Pre-emptible VMs
  • Go 1.8 on App Engine
  • Cheaper GPUs
  • Kubernetes

What were your personal highlights for 2017?

• Mark
  • Getting involved with SIG API Machinery with Kubernetes
• Melanie
  • Watching Haben Girma, the first Deafblind Graduate of Harvard Law School, speak about accessibility in tech.

It's the end of the year! So we'll be taking a break, and returning in January 2018!

This week we get the band back together! Francesc Campoy Flores rejoins the show

Deep Kapadia and JP Robinson from New York Times join Mark and Francesc to discuss how they use Google Cloud Platform to serve the New York Times to its readers.

JP Robinson maintains NYT's internal and open source tools and frameworks that are related to the Go programming language. He also lead backend development of NYT's games platform. Recently his team completely rewrote our backend with Go and GCP tools. In doing so they've managed to significantly lower request latencies and cut costs in half.

Deep Kapadia manages the Infrastructure and Delivery Engineering, Site Reliability and Test Automation teams at The New York Times. His teams are responsible for providing other engineering teams with tools and processes needed to get their jobs done on a day to day basis. His teams recently have been working on building the GKE deployment pipeline and enabling other teams to migrate to the Cloud from our physical datacenters and also moving their entire edge and routing caching architecture from internally hosted varnish to Fastly. They also helped move most of their site behind HTTPS.

• Cutting cluster management fees on Google Kubernetes Engine blog
• Coming in 2018: GCP's Hong Kong region blog
• Introducing an easy way to deploy containers on Google Compute Engine virtual machines blog

• New York Times Crossword site
• Moving The New York Times Games Platform to Google App Engine blog
• New York Times in 1996 webarchive
• Google App Engine site docs
• Cloud Datastore site docs
• Kubernetes Engine site docs
• Cloud Pub/Sub site docs
• Google BigQuery site docs
• Cloud Endpoints site docs
• Drone GAE github
• Drone GKE github
• Marvin github
• openapi2proto github
• gRPC site
• New York Time Open site

What best practices are there for securing a Kubernetes Engine Cluster?

• Precious cargo: Securing containers with Kubernetes Engine 1.8 blog

Mark will be Montreal in December to speak at Montreal International Games Summit.

Melanie will be at NIPS (Neural Information Processing Systems) in Long Beach and will also be attending Black in AI on December 8th.

Deep Kapadia and JP Robinson from New York Times

Myles Borins talks with Mark and Francesc about Node.js from its history, how to contribute, the consensus-seeking governance, and why it's important to Google Cloud Platform.

Node.js is an open-source, JavaScript runtime environment built on Chrome's V8 JavaScript engine, and Google is a Platinum Member of the Node.js Foundation.

Myles Borins is a developer, musician, artist, and maker he works for Google as a developer advocate serving the Node.js ecosystem he graduated with a Master of Music Science and Technology from c.c.r.m.a. a.k.a the center for computer research in music and acoustics

• Reduced GPU prices on GCP and preemptible local SSDs blog
• Skylake processors now available in 7 regions blog
• New Episodes of Learn TensorFlow and Deep Learning, without a PhD:
  • Modern Convolutional Neural Nets video
  • Modern RNN Architectures video
  • Deep Reinforcement Leanring video

• Node.js site
• Node.js on Google Cloud Platform site docs
• Node School site
• Node.js with Justin Beckwith podcast
• App Engine site docs
• Cloud Functions site docs
• Kubernetes Engine site doc
• Introduction to Kubernetes: How to Deploy a Node.js Docker App site
• Socket.io site

How do you give public postmordems?

• Fearless shared postmortems - CRE life lessons blog

Mark will be Montreal in December to speak at Montreal International Games Summit. Melanie will be at SOCML (Self-Organizing Conference on Machine Learning) end of this week and NIPS (Neural Information Processing Systems) in Long Beach next week.