Petr Romanov

• Created server logic and user interface for a desktop application managing document workflows, enhancing usability and performance. Additionally, improved team efficiency by developing installation scripts, reducing the setup time for new developer from three days to just a few hours. • Built resources to streamline document processing, significantly improving workflow efficiency for business applications.

• Developed virtualization software for the C167CR Infineon chip using C and C++ on the ARM 32 platform, successfully simulating the C167 processor to ensure software compatibility. • Designed and implemented a real-time data collection system in C++ and Qt for industrial devices on a Raspberry Pi 3 running Linux, enhancing data accessibility and monitoring capabilities.

• Helped develop a navigation system for university buildings using Apache Kafka, collaborating with a student team. • Created functional components and automated key processes for better system efficiency. • Learned advanced Shell and Python scripting and how to configure Java frameworks via config files. • Gained hands-on experience working with real-world Java systems and complex integrations.

• Led a team in the development of an augmented reality application that allows users to design and furnish virtual homes with customizable content, enhancing user engagement and creativity. • Successfully pitched the project and secured its selection for a prestigious technology exhibition, demonstrating its innovative potential. • Mastered JavaScript and custom AR frameworks while collaborating with international teams from the USA and Singapore, fostering cross-cultural teamwork and technical expertise.

Thesis Project: Automatic Data Redistribution for Adaptive Resource Management on HPC Systems • Developed a scalable system for flexible data partitioning and flow control across parallel machines using C, MPI, and HPC Chair’s libraries.

Thesis Project: Subsystem of GUI Elements Recognition • Developed a new subsystem for the "Sikuli" tool, extending its functionality to recognize specific GUI elements regardless of transformations or relative positions, using C++, OpenCV, Qt, and Python.