Blending Event Stream Processing with Machine Learning using the Kafka ecosystem


Details
Hey Kafkateers!
Se anche voi vi siete lasciati le ferie alle spalle e avete bisogno di un motivo in più per affrontare il rientro, vi aspettiamo al quinto Kafka meetup milanese: ci vediamo il 19 settembre in PoliHub (Via Durando 39, Milano, raggiungibile a piedi dalla stazione del passante Milano Bovisa), per un viaggio tra sistemi streaming ed algoritmi di machine learning, guidato dai ragazzi di Radicalbit... Il tutto in salsa Kafka, ovviamente. ;)
Agenda e ulteriori dettagli qui sotto, bloccate i calendari!
P.S.: il talk sarà tenuto in italiano.
-----
Agenda
6:45pm - 7:15pm: Accoglienza, salatini e birre, gentilmente offerti da Radicalbit!
7:15pm - 8:00pm: Talk: "Blending Event Stream Processing with Machine Learning using the Kafka ecosystem", Roberto Bentivoglio & Andrea Spina (Radicalbit)
8:00pm - 9:00pm: Domande e networking
-----
Talk
Title: "Blending Event Stream Processing with Machine Learning using the Kafka ecosystem"
Abstract:
During the Radicalbit Natural Analytics platform development the highest focus has been addressed to the integration of novels Artificial Intelligence capabilities to real-time based technologies, attempting to tackle the wide and challenging world of Streaming Machine Learning. Given that Kafka is a first citizen of the Radicalbit product, lately, the highest effort has been put on enriching it with machine learning features, the core point of RNA.
This talk attempts to share what has been achieved by Radicalbit by the R&D team and student trainees in these months: after explaining the main ML concepts and which changes Online ML brings, the speakers will justify the carried on work around this topic focusing on RNA and the Kappa Architecture idea; then, they will introduce former concluded projects at Radicalbit and their current work on Kafka by introducing two undertakings: k2o-serving, a module achieving streaming h2o model serving with Kafka Streams, and KafkaStreamsOML, a suite of online machine learning algorithms implemented on top of Kafka.
Speaker #1 - Roberto Bentivoglio
Roberto holds an MSc in Computer Science. He is studying, playing and experimenting with Scala since 2009 and he is using Scala for his everyday work since 2013. He spent four years in Switzerland where he worked, among others companies, for Credit Suisse Group. He worked until the middle of 2015 in DATABIZ as Head of IT Innovation and after that, he started to work in Radicalbit as CTO focusing on Fast Data technologies and in particular on Apache Flink and Apache Kafka. His main responsibilities are the evaluation of new technologies, the ideation and the delivery of the Radicalbit products.
Speaker #2 - Andrea Spina
Andrea graduated as Computer Engineer at "Università degli Studi di Modena e Reggio Emilia" (Italy), working on his master thesis at DIMA Group, TU Berlin, Germany and, along the same period, he was the co-author of "Benchmarking Data Flow Systems for Scalable Machine Learning" science paper. Andrea is currently working as a Data Engineer at Radicalbit, Milan. His main works have been focused streaming technologies, machine learning and security features.
-----
Se hai voglia di presentare o ospitare il nostro prossimo evento, non esitare a contattarmi: matteo@bottega52.it
Ti aspettiamo sullo Slack della nostra community (https://slackpass.io/confluentcommunity)!

Blending Event Stream Processing with Machine Learning using the Kafka ecosystem