Импортозамещение: Оценка производительности и совместимости Китайских GPU, при работе с различными SOTA ML моделями. Адаптация Китайского программного обеспечения для работы в РФ.

• Анализ данных по основной деятельности предприятия в информационных системах, вы- явление закономерностей; • Обучение нейронных сетей и построение предсказательных моделей по прогнозированию отцепов на ремонт, времени прибытия на конечный пункт, наличию свободных контейне- ров на терминале и т.д.; • Анализ и апробация новых идей и технологий; • Подготовка моделей для внедрения в сервисы Компании.

Используя Physics Based Deep Learning (PBDL) создавать модели с высокой предсказательной способностью для промышленности Швеции. Такой подход позволяет приблизить точность моделей машинного обучения к физическим. Разработка от этапа формулирования уточняющих вопросов к заказчику, до этапа отчёта перед заказчиком в виде готовых моделей и презентационных интерактивных страниц для наглядной демонстрации результата.

Административные Совершенствование программного продукта: добавление новых возможностей, доработка старых

Deploying the environment by Docker, Docker-compose, virtualenv, git. Exploratory data analysis (EDA). Formulation and testing of hypotheses. Development of parts of complex applications by Python. NLP, NER, GAN, time series. Main applications in: mining and processing plants, medical institutions, news resources.

All about my current work is under Non-Disclosure Agreement. All, except that I proud of working with so wise people:-)

R&D

Programmer-technologist

+ developed and implemented a series of experiments that allowed for the first time to discover the effects: 1. magnetomigration in solid-state nanosystems 2. magnetomigration during the deposition of granular magneto-heterogeneous films 3. GMR oscillation dependence from the thickness of granular cobalt-copper films. + made hydrodynamic's and moleculardynamic's model of magnetomigration in solid-state magnetic nanosystems. + the method of controlling the surface morphology of Co-Cu films obtained by ion-plasma spraying has been developed. + the method of obtaining magnon crystals using selective ion-plasma deposition and the effect of magnetomigration. + the certificate of Rospatent on state registration of the computer program 2014619177 was received In YAF URAN FTIAN performed researched materials for: 1. MRAM (Magnetoresistive Random Access Memory) 2. Magnetoresistive sensors 3. Solid-state magnetic nanosystems. For what handled: 1. Deposition of thin films by ion-plasma spraying, and their study. 2. Development of an automated stand for the study of magnetoresistance. 3. Molecular dynamic simulation using the LAMMPS package. 4. Preparation of research landscapes to improve the efficiency of experimental studies. 5. Actively participated in the organization and administrative support of research. Was elected Deputy Chairman of the Yaroslavl Council of Young Scientists of the Institute. The results of research activities were awarded with a number of awards, such as a diploma for winning the competition of the best reports of the international youth conference "The way to science" YarSU, 2014. "Leadership qualities, ability to work effectively in a team, good intellectual potential, active citizenship, ability to learn and develop new activities, as well as high motivation for professional achievements and development" are noted in the recommendation letter from the Chairman of the Supreme Council of the Russian political party "Edinaya Russia" B. V. Gryzlov.

Delphi: Databases + GUI Data security.

Postgraduate, Condensed state physics

Физический, Микроэлектроника и полупроводниковые приборы