On Saturday, April 27th, this years Global Azure Bootcamp (GAB) took place. It’s been the 7th GAB and I attended for the 3rd time – and as part of my contribution to the community also did a talk on Azure Machine Learning Studio. But that is a story I will tell you more about later in this article.
The location – Microsoft Sales Office in Hamburg-Bahrenfeld
This year the GAB happened in the Microsoft Office in Hamburg-Bahrenfeld. I like this location as you can get there really easy by using the nearby highway departure and dont need to drive through Hamburg city traffic. I don’t care for the additional kilometers. On the other hand one of the major disadvantages of this location is that it’s really small. If you had about 90 participants as planned for this years bootcamp it’s a really small space and the rooms, air condition and some other facilities in the building are obviously not made for conferences and meetups at this size.
In #Hamburg startet das #GAB2019 mit der Keynote von @JanDamaschke – voller Raum und neben mir stehen noch Leute… Ich wünsche allen Teilnehmern einen angenehmen und lehrreichen Tag!#GlobalAzure #azure #MVPBuzz #CommunityRocks #communitypower #KuC #azuremeetuphh pic.twitter.com/cXHgx1dd47
— Bjoern Peters [MVP] (@SQL_aus_HH) 27. April 2019
This became pretty obvious already at the key note by Jan-Henrick Damaschke. As you can read from a tweet by Björn Peters (MVP, Data Platform) the room was full and a lot of attendees needed to stand around in the background to join the session.
Two different tracks, “session” and “workshop”
This is why I liked that they planned the agenda with 2 different tracks, called session and workshop. The session track provided overview-level sessions which introduced attendees to common Azure key concepts and options.
This is the track where they discussed several aspects of cloud security, like threat protection, Firewalls as a service, and update management in the cloud, as well as containers, data portability and management and (my talk) a session on artificial intelligence and machine learning.
I love that type of sessions where you can get good background knowledge about key concepts and options as it is a frequent requirement by my customers to get recommendations and insights for how to plan their solution design and get more in-depth knowledge of how things are going on a move to the cloud.
Nevertheless I joined the PowerApps loves Azure Functions workshop in the morning. While I think I’m pretty familiar with the concepts of PowerApps and Microsoft Flow, I feel like I need more insights to PowerApps and delimination between a PowerApp and a “classic” Web or Mobile App – and how to interact with your Web App and/or Rest API.
First Workshop at #GlobalAzure Bootcamp #hamburg with @TKarafilov and @PapaRiedel starting now! #PowerApps and #Azure integration incoming 😉#gab2019 pic.twitter.com/vaK4t8TcqG
— Jan-Henrik Damaschke (@JanDamaschke) 27. April 2019
PowerApps loves Azure Functions workshop
On the other hand there was the workshop track with longer sessions to get more hands-on experience with the look and feel of a cloud service. The workshop in the morning discussed how to get serverless with some basic apps in Azure.
In their example, Stefan Riefel (@papariedel) and Tomislav Karafilov (@TKarafilov) showed how to get started with your Azure Subscription and how to activate this either with your Microsoft Account or with your existing Azure AD Account from your companies environment.
In a second step, Stefan showed how to start with a simple PowerApps canvas app that users may create as a citizen development. This app was then used to embed and run several on-click activities to launch Azure functions using different parameters.
In a third step, Tom focussed on how to build a simple Azure function V2. While the function itself was basic, remember we are in a session for beginners, he really did a good introduction in how you make use of parameter bindings in your functions and create a swagger API file from the Azure functions template. They showed how to customize the template, download it and import it to your PowerApp.
Finally, Stefan showed how you can put that all together to run your Azure function form the PowerApps button.
Really great stuff! So many options, so little time. I feel like now I’m good to go and join Peter Wiens session on PowerApps 101 at the Office 365 meetup in August, 2019. And I will create a more responsive UI for some of my web services now that I know how to connect my custom Azure API with a PowerApp
Lunch break – time to say thank you to the sponsors
This GABs lunch package is sponsored by Subway and Microsoft. So I took a really tasty Subway to go package and checked how the other GABs in Germany and DACH area were going. In Hamburg, I’m glad that the no-show rate was really low, this time. Unfortunately this was not the case on some other locations.
When you hear things like that, you really should take a look on who did not show up, I feel.
Yes, there may be reasons why you cannot attend to an event you registered before. No, there is no reason not to click “I will not join”, send a mail or choose the other appropriate option to let the organizaers know how many people will really join. This not only helps the organizers to use the event resources but also save a lot of money when it comes to rooms size and food packages, which still need to get paid.
In Hamburg, about 75 percent of meetup attendees are consultants themselves. If you are a customer or end user, I kindly like to ask you that if you see a lot of no shows for your event, please go back to the meetup sign-up page and check the list of attendees. Check for the consultants and service providers. Imagine, how they may care for your requirements in times were life is not going the easy way.
Beeing a no show may also give you a hint on how reliable this consultant for your business need?
What do you think?
Weiter gehts- Firewall as a Service #AzureGlobal #Azure #KuC pic.twitter.com/CkOZLKoG0e
— KuC/Wolfgang Wündsch (@KWundsch) 27. April 2019
My first session in the afternoon discussed usage of the Firewall as a service offer by Microsoft compared to other solutions, like NSGs, Application proxies (load balancer + WAF) and 3rd party firewall products like Checkpoint, Fortinet, Cisco and others. Really impressed about the do’s and don’t that came up during that sesison.
Müssen wir SQL on Azure eigentlich sichern? #AzureGlobal #Azure #KuC pic.twitter.com/Eu3xosrXk6
— KuC/Wolfgang Wündsch (@KWundsch) 27. April 2019
After that I needed to get back to the speakers corner to prepare my own session, so unfortunately I missed Fabian Baders talk on update management with Azure. But luckily I went finished before Björn Peters started his session on backup and restore for SQL services on Azure. He did a good overview of the different options when using Azure SQL Database, Elastic Pools and Managed Instance and did a comparison to what’s possible when you ran your classic SQL Server virtual machine, doesn’t matter if you use the marketplace image for deployment or if you installed your bring-your-own-license to an existing VM.
Machine Learnings – Session startet #AzureGlobal #Azure #KuC pic.twitter.com/ogHXvDYgYw
— KuC/Wolfgang Wündsch (@KWundsch) 27. April 2019
After Björn finished, it was time for my own talk. Azure Machine Learning – a real business use case. What do you feel about a machine learning session starting in the later afternoon? You’re right! Now that people are feeded and awaken you should really let them dive in into data quality and requirements that come with normal distribution so they know what to check the customers dataset for: skewness, density, validity and some more important things when it comes to stochastic calculus and how that impacts the quality of your prediction. Doesn’t sound like an Azure Machine Learning session? You’re wrong. From my experience as a consultant I feel that most customers who try play around with AI and machine learning stumble not for the reason that the tools are too complicated but as they dont undersand the wording for the defferent tasks as used in Machine learning studio and all the mathematical background.
Now that we got a good overview about how prediction works we dive deeper into Machine Laerning studio itself. If you see the prologue clear, it’s really a simple task to create your training experiment. Of course, an intro-level talk does not get a complex example. So my experiment showed how to import your dataset, calculate string-columns to numeric by using scored hashes and finally train and export your model.
Once you get your trained model, we create our web service to use this with an Excel addin or from your custom app by calling the URL with parameters.
To be honest, we didn’t dive into the custom API calling to deep as time flies by and the organizers gave a hint for the time left. So we did a short lessons learned and review of my personal experience when it comes to Azure Machine Learning studio.
Please excuse me that I just did a summary that sounds more like a teaser for that talk on Azure Machine Learning studio, but maybe readers plan to join my talk at SQL Saturday #856 Rheinland; as I plan to use the same business use-case there, I don’t dive in the use-case in this article.
Speaker Corner @ #GlobalAzure #Hamburg@_burbert @AtholyMardies @jandamaschke pic.twitter.com/R8EBaRiwOQ
— Fabian Bader (@fabian_bader) 27. April 2019
It’s always a great pleasure to join events like the GAB in Hamburg if you also do a talk so you can go to the speakers room and see all these fantastic people who provide you with so much technical input on possibilities, try to inspire you on how to use the one or the other solution and answer questions you though you never will need to ask. Yes, using a cloud service sometimes raises questions you never thought you will need to consider.
Sharing means caring and this is especially true for a community event. While some people do this for marketing and reputation, it’s great to see many of the speakers who really act to support and enable other attendees. For example, Christoph Burmeister and I had a great chat on different options for home automation.
In this conversation, we discussed different options like intelligent light bulbs, automation by using geofencing options and more. Some of you may also know that I personally prefer (some older) Google Home devices where I developed some capabilities that are connected to my Azure web app services and control fresh water tank pump out and fill-up using Windows 10 IoT devices with some sensors.
Ah’, and finally let’s not forget that there was thate adorable GAB2019 sticker which now decorates my laptops back.
And now that GAB 2019 is gone…
If you’re interested in public cloud computing and Azure I hope you had an opportunity to also visit a Bootcamp near to your place. For them who live near Hamburg I also like to advice that there is a monthly opportunity to join the local Azure Meetup Hamburg, please see https://www.meetup.com/de-DE/Azure-Meetup-Hamburg/?chapter_analytics_code=UA-97255023-1 to join yourt local community.
As an attendee of my session you may also be interested in the local AI meetup or Azure and you want to join Microsoft AI User Group Hamburg to get deeper insights. If so, please join them at https://www.meetup.com/de-DE/Microsoft-AI-User-Group-Hamburg/ , but you need to hurry up as their next meetup is on Monday, April, 29.
Anyway I hope you got some inspiration on what is possible with the Microsoft cloud and how to get things done. If you liked GAB 2019, I appreciate to see you at one of the next local meetups here in Hamburg or on next years GAB 2020!