Maciej Próchniak
From Touk
More than 10 years of experience, currently usually as architect/(lead) developer - but my roles vary from analisys to deployment. Most of it Java & Scala, also some Groovy experience.
My main fields of interest are integration (Camel, OSGi) and stream processing systems (Akka, Kafka, Flink) but I also did my share of "enterprise" systems with Spring, Hibernate and so on.
I also like to give talks at conferences - Confitura, JEEConf, VoxxedDays just to name a few.
Currently I’m leader of TouK Nussknacker project which enables analysts to create streaming jobs with friendly UI.
Blog: mproch.blogspot.com
Architecture
“Zero code” systems - hopes, myths and reality
Business owners always dream about system that will let them change its behaviour without involving coders. This desire is event stronger in enterprises, when coders are usually in different company. We’ve seen many examples: BPMN, DSLs, Drools, and many in-house samples. Big vendors and open source projects promise miracles - and usually do not deliver - or at least not to expected extent...
During more than 10 years in our industry I’ve seen and taken part in many such efforts. I’d like to share my experiences - what is (almost) doomed to be a failure, what can only depend on structure of organisation and when can we expect some modest success.
I also believe that the project I’m working on - Nussknacker (check it out on github) can be a bit more successful. Join me to hear about our domain (real time stream processing in telco), technologies (Kafka, Flink, React and more) our vision and architecture.
Big Data & AI
Stream processing in telco with Apache Flink & TouK Nussknacker
Two years ago we helped to introduce Apache Flink in one of the biggest mobile operators in Poland - at first to help with real time marketing. The data used included information from billing and signalling systems. We wanted to enable analysts and semi-technical people to create and monitor processes and that’s how Nussknacker - our open source GUI for Flink was born.
Today we’re processing billions of events daily from more than dozen of sources with > 40 processes running in production. Most of those jobs are created by analysts, without the need of developers assistance. We use our cluster not only for marketing but also fraud detection.
I'll show how we achieved that - I'll briefly describe our use case and architecture, and then show short demo of Nussknacker.