2018 Firebase Summit Prague

kamil

At the end of October, I had the opportunity to visit Prague to attend Firebase Summit – a conference centred around Firebase, Google’s developer platform. This platform provides a great toolset for building high-quality mobile apps and helping businesses grow. We’ve used Firebase a lot this year and believe that it’s a great solution in many situations. We were really excited to hear about the new advancements they have been making and this could influence our products!

The conference this year focused all around smarter data analytics and growing your user base. Google announced multiple improvements to products such as Analytics, Predictions and Remote Config, providing the ability to join data from multiple different sources. It is now possible to export Firebase data to BigQuery – Google’s tool for performing deeper data analysis, to evaluate certain patterns and potential issues affecting your business. Google also published a new case study of their customer, Hotstar, about how they managed to increase the user engagement by 38%.

That all sounds great, but how that will affect my existing app? Well, it’s now possible to re-engage users by sending push notifications targeted specifically to those who have used the app in the last few days, or those who haven’t triggered a certain analytics event in the app, such as the registration flow. Segmentation tools have been greatly improved to provide more flexibility over who we can target and means we can avoid spamming other users. The process for long running campaigns has been simplified as well – it is now possible to create a campaign once that will continue to trigger until stopped. These and many other examples were mentioned during this talk.

However, the most interesting talk for me was the showcase of Google’s Machine Learning tools both in Firebase Predictions and in mobile ML kit. With a single click on the dashboard, Firebase Predictions will start to analyse the dataset you have gathered through Analytics and provide you with insightful knowledge, such as the users’ churn rate or whether or not the user will be likely to make a purchase within the app. This tool, teamed with Remote Config, will allow us to perform actions that were not possible before. For example, we will be able to dynamically display ads within your app based on whether the user is likely to make an in-app purchase or not. All of this works without any extra development effort and uses Google’s battle-tested ML algorithms, so it’s likely to be very accurate.

Machine Learning can not only be used for data analysis, but also for certain use cases within your app, without the need for any internet connection. Firebase provides built-in models for analysing human faces, recognising text, scanning barcodes and recognising landmarks. It can also be used with your own TensorFlow Light models, meaning that it can cover custom needs, such as the ability to recognise country flags.

The very last announcement worth mentioning is a development improvement to Firestore database – the ability to run the database emulator locally. This is a big improvement for developers which will greatly improve the testing flow and efficiency of development.

Overall, the Firebase Summit was a fantastic learning opportunity for us. At GearedApp we are always looking to stay ahead of the curve and we are looking forward to putting some of these new features into practice. While I was at the Summit I also had a chance to meet many great people from the Firebase team and other conference attendees. We shared ideas, brainstormed on how to fix certain problems, played games, took loads of pictures and had fun overall. These amazing people are what makes this conference so great and I’m glad I had the opportunity to take part in it this year.

Last but not least, here’s a portrait of me, Firebase style: