Олександр

Reshetar Oleksandr | Curriculum Vitae
ć [відкрити контакти](див. вище в блоці «контактна інформація») • Ukraine, Kyiv • × [відкрити контакти](див. вище в блоці «контактна інформація»)
a GitHub • ] LinkedIn

Summary
Bachelor’s student in Applied Math specialization. Constant passion for seeking the possibility
of improving hard and soft skills. Enjoy participating in projects with different technologies to
widen my knowledge area.

Education
NaUKMA: National University of Kyiv-Mohyla academy 09/2019 - now
Applied math
During my study, I have completed several fundamental math courses, like discrete math,
analysis, linear algebra, etc. The knowledge of linear algebra gave me an opportunity to
comprehend the logic behind machine learning algorithms, where vectors and matrices
are multiplied. The graph theory gave me a good background in solving DFS/BFS
problems.

Experience
JUNIOR/MIDDLE SOFTWARE ENGINEER AT EPAM SYSTEMS 01.08.2021-NOW
During my software development I was deeply involved in designing and creating API
for online shops, e.g. e-commerce. Our client Kingfisher has several distributions in
France, Poland, and UK. As a framework, we use ATG, which is the predecessor of Spring
Framework. My tasks included:
• Developing new API calls

• Bug fixing
• Taking part in Sprint planning and estimating the user stories
• Writing scripts for adding new products into the database

• Taking part in Code Review
• Unit testing
• Automatization testing

External courses at EPAM Systems 11/2020 - 3/2021
The courses covered the fundamentals of Java. I have learned about Stream API,
Concurrency, REST API, and Spring framework. As the final task, I had to develop one
project using two technologies: Spring Framework and Servlet API. My responsibility
was to design and write the ”Selection Committee” application for enrolling students at
University. To manage the system with GUI, I have used Thymeleaf Engine. Unlike the .html
files in the Spring project, the Servlet app required the usage of JSP pages. The endpoints
were secured with Spring Security, so only users with privileged roles can change the
state of the system. For storing the required data I have configured the MySQL Server. In
addition, the Servlet app demanded the implementation of the DAO pattern to save and
get the objects to/from the database properly. Both projects included ”Internationalization
i18n”. This experience gave me an insight into how the REST API applications work under
the hood.

1/2
Skills
Programming Language Middle: Java Junior: Python | JS | Haskell
Frameworks & Tools Git | Gitlab | Gradle | Postman | Maven | Docker | Jira | Confluence | Jenkins | Intellij
IDEA | Oracle DB | Oracle SQL Developer | Hyper-V Manager | Spring(boot, jpa, web)
| Kubernetes | Kafka| RabbitMQ | Log4J | Mockito | Swagger
Libraries Junit | numpy | opencv
Languages Native: Ukrainian | Middle: English B1+ | German B2

Extracurricular Activities
Volunteer 4/2022 - 5/2022
Student organisation You4Ukraine
I was involved in developing a Merch shop for Ukraine support. The back-end was written
using the Spring framework. My task was to implement API for managing items in the shop.
Additionally, I had to implement a mapper mechanism in order to convert the db object
into a response object(dto) and vice versa. I was also responsible for adding a security
layer, such as Spring Security. To deploy the application we have used AWS EC2 instance.
AWS S3 stored static data, like images. The front-end was written on React and is available
here

ML Course 10/2022 - 12/2022
CSC2431 : Artificial Intelligence in Medicine University of Toronto
During my participation in this course I was doing in pair the research on Coreset-Selection
algorithm: that is, how to reduce the size of the original dataset with minimal loss of
accuracy. I have investigated different coreset-selection methods, such as
• Uncertainty Based Method

• Error/Loss Based Methods
– Forgetting Events
– GraNd and EL2N Scores
– Importance Sampling

All experiments were tested on the GPU cluster provided by Vector institute. The results
of the experiments were described in final report

NASA Hackathon 10/2022 - 10/2022
Nasa Spaceapp Challenge
During the hackathon, my team was involved in training an existing machine-learning
model in order to detect enemy vehicles. Due to restricted time, we have found a dataset
containing only videos and objects of one class. As a baseline model, we have chosen the
YOLOv5. My task was to prepare a dataset for training: parsing video on frames(images),
and converting JSON metadata into txt files with bb coordinates. As a programming
language, I have chosen Python(the application was running on google colab), since it
suits the best for small script files. Additionally, we have created a Telegram Bot, so that
users can send a photo and obtain back it with the annotated object.

2/2