Artem Zavalko
About Artem Zavalko
Artem Zavalko is a Software Architect with extensive experience in agile software development, particularly in the banking and finance sector. He has worked at various companies, including DataArt and Finanz Informatik Solutions Plus GmbH, and has contributed to projects focused on handwriting recognition technology.
Work at Finanz Informatik Solutions Plus
Artem Zavalko has been employed at Finanz Informatik Solutions Plus GmbH as a Software Architekt since 2013. He has accumulated over 11 years of experience in this role, contributing to various software architecture projects in Frankfurt am Main, Hesse, Germany. His responsibilities include overseeing technical aspects of software development and ensuring alignment with industry standards.
Previous Experience at DataArt
Prior to his current role, Artem Zavalko worked at DataArt as a Senior Software Entwickler from 2007 to 2011. During his four years in St. Petersburg, he gained valuable experience in software development, which laid the foundation for his future roles in software architecture.
Experience at Pixell, The Amadeus Leisure Group
From 2011 to 2013, Artem Zavalko served as a Software Architekt and Senior Software Entwickler at Pixell, The Amadeus Leisure Group, located in the Bonn Area, Germany. His two-year tenure involved significant contributions to software architecture and development, enhancing his expertise in the field.
Education and Expertise
Artem Zavalko studied at ITMO University from 2007 to 2010, where he earned a Dipl.-Ing. in Informationssysteme und Technologien. He also attended Bratsk Staatliche Universität from 2003 to 2007, focusing on the same field. His educational background supports his extensive experience in agile software development, particularly within the banking and finance sector.
Projects and Contributions
Artem Zavalko has been involved in several notable projects, including efforts to reduce handwriting recognition error rates to 2%. He co-authored an article on handwriting recognition using a Convolutional Neural Network (CNN) and experimented with JavaCNN and Deeplearning4J for neural network configurations. His work includes optimizing a CNN to achieve a 98% accuracy rate in recognizing handwritten digits and developing a solution for the automatic recognition of handwritten tax ID numbers.