Machine Learning As A Service : Setting up a decoupled, streaming architecture to leverage the power of ML models
These days developers have the power of Machine Learning libraries at their disposal. Many of us experiment with these tools in our daily work, but do not manage to scale the results or reuse the stack outside of our team.In this talk you will learn how an enterprise like ING Bank built a decoupled, streaming Machine Learning service to enable all IT teams to use their data to leverage the power of ML. In a live demo, we show how the platform is able to classify incoming user feedback using Kafka and Docker containers to enable teams to instantly act on current sentiment of their product.
With more than 20 years of experience developing software for the web, Effi starting out as a multimedia specialist. Later he switched to backend technologies. He loves to learn and share new technologies and paradigms. Organizing and contributing to coding dojo's, knowledge sessions, blended learning trainings, conferences/meetups, and giving trainings, there is one value he holds highest: Sharing is caring. This includes brewing and sharing beer in his free time. Currently he is working for ING, creating a Machine Learning Model serving platform (MLaaS).
I'm a passionate software engineer with a balanced lifestyle. That means a good dose of Java, Web Components and everything in-between during daytime with a second helping of low-level programming and machine learning at night.