Skip to content
View b1zmark1's full-sized avatar

Block or report b1zmark1

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
b1zmark1/README.md

Nikita Patrushev

Machine learning and data systems for chemical engineering, industrial processes and document intelligence.

Email | Telegram | GitHub | GitLab | Saint Petersburg

Python Machine Learning Data Science Computer Vision LLM Docker

About

I work at the intersection of chemical engineering, industrial data analysis and machine learning. My projects focus on practical ML systems: forecasting process parameters, analyzing industrial datasets, building OCR/LLM pipelines and turning research ideas into usable prototypes.

I am especially interested in applied AI for chemical production, oil and gas, process monitoring and engineering decision support.

Selected Work

Project Area What it does
NeftecodeTeamRocket Industrial ML Predicts oxidative test results for multi-component oil formulations. Neftecode 2026 winner project.
SuccessfulBPMN OCR / LLM / System Design Asynchronous ML service for scanning and generating BPMN diagrams with detection, OCR and LLM-based structure recovery.
CoalGuard Time Series / Monitoring Demo platform for coal pile monitoring and 3-day fire risk forecasting.
DistillationColumnTemperaturePrediction Chemical Process ML Forecasts distillation column top temperature using weather data and regression models.
SteinerRL Reinforcement Learning Grid-based Steiner tree pipeline for pipeline network design experiments.
CorrProgram Data Analysis Tooling App for correlation analysis from CSV/Excel data with visualization and export.

Highlights

  • Winner of Neftecode 2026 with an ML solution for oil formulation analysis.
  • Prize winner at ITMO MegaSchool, ML System Design track.
  • Prize winner of a Gazprom Neft olympiad with a full-stack ML concept for distillation column monitoring.
  • Speaker and author on ML applications in chemical technology and industrial process forecasting.

Tech I Use

Python | pandas | scikit-learn | PyTorch | time series | computer vision | OCR | LLM pipelines | FastAPI | Streamlit | Docker | JavaScript

Current Direction

I am building stronger end-to-end ML systems for industrial use cases: from data preparation and model training to backend services, interfaces, deployment and clear project documentation.

Pinned Loading

  1. SuccessfulBPMN SuccessfulBPMN Public

    SuccessfulBPMN — асинхронный ML-сервис сканирования и генерации BPMN диаграмм

    JavaScript

  2. NeftecodeTeamRocket NeftecodeTeamRocket Public

    Модель предсказания результатов окислительного теста Даймлера

    Python 1

  3. CoalGuard CoalGuard Public

    Демо-платформа для мониторинга штабелей угля, прогноза риска возгорания на горизонте 3 суток и просмотра фактических воспламенений.

    Python

  4. SteinerRL SteinerRL Public

    Grid-based Steiner Tree reinforcement learning pipeline with mask-to-graph conversion, A* baseline, Gym-like environment, GCN policy, and PPO training.

    Python

  5. DistillationColumnTemperaturePrediction DistillationColumnTemperaturePrediction Public

    В репозитории показан способ прогнозирования температуры верха ректификационной колонны (колонна ДИП) с использованием метеоусловий города Москва и регрессионной модели машинного обучения.

    Python

  6. CorrProgram CorrProgram Public

    Python app for calculating correlation matrices from CSV/Excel data with Streamlit/Tkinter UI, Plotly heatmaps, p-values, missing-value handling, and export to XLSX, HTML, PNG, and PDF.

    Python