For one of the closing keynotes of DevRelCon Earth 2020, I spoke about what I call the Developer Advocacy Feedback Loop. People often think about developer relations and advocacy as just being about external outreach. However, there’s more to it! Developer Advocates are here to represent users, developers, technical practitioners, to influence the roadmap and development of the services and products to suit their needs. That’s the internal advocacy that loops back into improving the products.

For the 24 hours of talks by Google Cloud DevRel, I recorded my talk about the new Java 11 runtime for Google Cloud Functions. I wrote about this runtime in this article  showing for example how to run Apache Groovy functions, and I also wrote about it on the  GCP blog  and Google Developers blog as well.

In this video, I’m giving a quick explanations on the serverless approach, the various serverless options provided by Google Cloud, and then I dive into the various shapes Java functions can take (HTTP and background functions), the interfaces you have to implement when authoring a function. And I also do various demonstrations, deploying Java functions, Groovy functions, or Micronaut functions!

With my partner in crime, Mete Atamel, we ran two editions of our “Pic-a-Daily” serverless workshop. It’s an online, hands-on, workshop, where developers get their hands on the the serverless products  provided by Google Cloud Platform:

• Cloud Functions — to develop and run functions, small units of logic glue, to react to events of your cloud projects and services
• App Engine — to deploy web apps, for web frontends, or API backends
• Cloud Run — to deploy and scale containerised services

You’ve certainly interacted with webhooks at some point: with a Github commit webhook, for Slack or Dialogflow chatbots, for being notified of Stripe payments, or when you receive an SMS via Twilio. The concept is fairly well known, but there are some roadblocks along the way, whether you implement a webhook handler (the URL being called) or a webhook backend (the service notifying URLs). It’s not necessarily as trivial as it may first seem. As I’ve been interested in Web APIs for a long time, I decided to look into this topic a bit more, by working on a new talk.

Last week, I was in Tokyo for the first time, to speak at the Google Cloud Next conference. During the DevDay, I spoke about Google App Engine and its 2nd generation runtimes, and I also presented Cloud Run on how to deploy and run containers in a serverless fashion. It’s been awesome to visit Japan for the first time and get a chance to meet developers there. Here are the slides I presented:

At DataXDay 2019, last week, I had the chance to present an updated version of my introductory talk on the serverless compute options on Google Cloud Platform. There’s always something new to cover!

For instance, if I put my Java Champion hat on, I’d like to mention that there are new runtimes for App Engine standard, like the beta for Java 11, and there’s twice the amount of memory as before. On Cloud Functions, we have an alpha for Java as well (currently Java 8, but it’ll be soon moved to Java 11 instead, as customers are more interested in the latest LTS version)

Last week at the Google Cloud Next conference, I had the chance to speak about the Java developer’s journey through the “serverless” offering of Google Cloud Platform, with my colleague Vinod Ramachandran (Product Manager on some of our serverless products):

Serverless Java in 2019 is going to be ubiquitous in your favorite cloud. Well, it’s actually been 10 years since you could take advantage of Java on Google App Engine. But now you can run your apps on the brand-new Java 11 runtime. Not only servlet-based apps but also executable JARs. And what about authoring functions? Until now, you could only use Node or Python, but today, Java is the third runtime available for Google Cloud Functions. We will review the various ways you can develop your Java functions. Last but not least, thanks to serverless containers, containerized Java workloads run serverlessly, without you caring for infrastructure, scaling, or paying for idle machines. Through various demos, we will look at the many ways Java developers will be able to write, build, test, and deploy code in Java on the rich serverless offering of Google Cloud Platform.

At GR8Conf Europe last year, I talked  about how to take advantage of the Google Cloud machine learning APIs  using Apache Groovy.

With Groovy, you can call the Vision API that recognises what’s in your pictures, or reads text.

You can invoke the Natural Language API to understand the structure of your text.

With the Speech-To-Text API, you can get transcriptions of what’s been said in an audio stream, or with Text-To-Spech, you can also generate human-like voices from your own text.

At Voxxed Days Microservices, in Paris, I talked about the latest development in serverless solutions on Google Cloud Platform, to deploy functions, apps  and even containers.

I answered an interview  on the theme of microservices, and how this maps to the Google cloud products.

And the video of my presentation was published on YouTube:

Here’s the abstract of the session:

In a matter of a few years, Google Cloud Platform has evolved from a very small set of products or APIs to a wealth of close to a hundred of products, services and APIs that developers can take advantage of.

This week, at the event Le Meilleur Dev de France, I gave an introduction to the whole platform, focusing on three key axis: compute, storage and machine learning. After an introduction on famous users of GCP, like Snapchat, Spotify or PokemonGo, I also gave a few examples of big French companies as well as French startups who have decided to go to the cloud with Google.