Responsibilities: - Supervising and assisting junior programmers (interns) in their work - Developing new software modules - Analyzing and optimizing code - Optimizing development processes - Leading backend development - Creating integration solutions to ensure high software performance - Designing system architecture - Communicating with clients and adjusting technical specifications for work Achievements: - Developed the TPT3 program for massively parallel Monte Carlo simulation of radiation passing through structural elements of nuclear facilities (see portfolio) - Porting the created software code to GPU using CUDA and OpenAcc resulted in more than an order of magnitude acceleration of complex physical modeling calculations - Implemented the Gitlab CI/CD system, which significantly accelerated all work processes related to code deployment and debugging - Organized the work of employees (interns) from scratch - Conducted several dozen code reviews - Actively participated in the development of several commercial projects necessary for creating digital twins of our company’s products - Was repeatedly invited to speak at seminars and thematic conferences as a speaker The latest GitHub repo containing the code of benchmarking sorting algorithms using parallelization on CPU and offloading on GPU using CUDA: https://github.com/Agar92/benchmarksortingalgorithms.git The GitHub repo with a sample project CURRENT_WORK/: https://github.com/Agar92/MyProjects.git The Phd thesis, thesis abstract and presentation are avalaible at: https://disk.yandex.ru/d/_PDekiB0AhM2kg Published versions of my articles: https://disk.yandex.ru/d/5Eht60GIPzFMAA

Responsibilities: - Writing, debugging and testing of a module for detecting the development of magnetohydrodynamic instabilities in thermonuclear plasma within a large software code for numerical simulation of plasma dynamics in a tokamak - Validation and verification of developed computational methods using experimental data Achievements: - Successfully achieved high-precision detection of magnetohydrodynamic instability development in the tokamak plasma simulation program - Managed to determine plasma optical density as well as its ionic and electron temperatures with an error of no more than 15% compared to experimental data, which is an excellent result for simulating thermonuclear processes