← Back to list
middle
Registration: 08.02.2024

Ivan Dobrosovestnov

Portfolio

Moscow Programming School

- Teaching a C++ and Python courses. - Development of tasks for the inner olympiad and C++ course.

SIT Programming School

Analysis of student’s programming code, improvement of the educational process through the integration of ML technologies: Developed a system that detects abnormal behavior of students, such as cheating, misunderstanding of the material (also known as “fraud detection”). - US Patent “Method and system for classification of student progress in solving a complex problemMethod and system for classification of student progress in solving a complex problem №18/175,551 (Filed Feb 28, 2023, on patent pendind). - US Patent “Method and system for automatically assigning a behavioral category to a student’s study” №18/175,551 (Filed Feb 28, 2023, on patent pendind). Developed an infrustructure for churn prediction analysis: - building ML pipeline from scratch (EDA / data collecting / prediction model / deployment); - NLP analysis of CRM tasks; - achieving 80% PR AUC score when predicting student’s churn. Developed Transformer Encoder-Decoder model for Knowledge Tracing. - Paper “Profiling of Students’ Competencies for Adaptive Learning Systems” (bachelor’s thesis).

Moscow Programming School

- Teaching a C++ and Python courses. - Development of tasks for the inner olympiad and C++ course.

Skills

Python
C++
C
Pytorch
Hadoop
Spark
Jango
Git
CI/CD
Docker
Ansible
SQL
DVC
Linux
Computer Vision
NLP

Work experience

Middle ML engineer
since 07.2022 - Till the present day |SIT Programming School
C++, Python
Analysis of student’s programming code, improvement of the educational process through the integration of ML technologies: Developed a system that detects abnormal behavior of students, such as cheating, misunderstanding of the material (also known as “fraud detection”). - US Patent “Method and system for classification of student progress in solving a complex problemMethod and system for classification of student progress in solving a complex problem №18/175,551 (Filed Feb 28, 2023, on patent pendind). - US Patent “Method and system for automatically assigning a behavioral category to a student’s study” №18/175,551 (Filed Feb 28, 2023, on patent pendind). Developed an infrustructure for churn prediction analysis: - building ML pipeline from scratch (EDA / data collecting / prediction model / deployment); - NLP analysis of CRM tasks; - achieving 80% PR AUC score when predicting student’s churn. Developed Transformer Encoder-Decoder model for Knowledge Tracing. - Paper “Profiling of Students’ Competencies for Adaptive Learning Systems” (bachelor’s thesis).
Teacher / Methodist
01.2020 - 07.2022 |Moscow Programming School
C++, Python
- Teaching a C++ and Python courses. - Development of tasks for the inner olympiad and C++ course.

Educational background

Applied mathematics and Information Science
2019 - 2023
Higher School of Economics University
Mathematics, Computer Science and Engineering
2017 - 2019
Higher School of Economics Lyceum

Languages

EnglishIntermediate