Skip to content
View LehaDeev's full-sized avatar

Block or report LehaDeev

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 supported. This note will be visible to only you.
Report abuse

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

Report abuse
LehaDeev/README.md

👋 Hi, I'm Aleksei Deev

Machine Learning Engineer · Data Scientist · Python Developer

Telegram LinkedIn Email


🧠 About Me

Data Scientist & ML Engineer with production experience in building end-to-end AI systems and a strong engineering background. I combine deep technical expertise in machine learning with real-world business acumen — having scaled an e-commerce business to ₽46M annual revenue and built commercial SaaS products from scratch.

  • 🔭 Currently working on ML systems, algorithmic trading bots, and AI-powered SaaS services
  • 🎓 Completed Yandex Practicum — Advanced Data Science Specialist (2026)
  • 💡 21 completed ML projects across classification, regression, NLP, CV, time series, and Big Data
  • 🏭 9+ years of engineering background in nuclear, oil & gas, and defense industries
  • 📍 Nizhny Novgorod, Russia · Open to remote work

🛠️ Tech Stack

Machine Learning & Data Science

Python Scikit-learn PyTorch TensorFlow Keras XGBoost LightGBM CatBoost

NLP & Computer Vision

HuggingFace BERT NLTK OpenCV ResNet

Data Engineering & MLOps

Apache Spark Docker FastAPI Redis PostgreSQL GitHub Actions

Data Analysis

Pandas NumPy Matplotlib SQL SHAP


🚀 Featured Projects

📈 we_trust_in_people — Algorithmic Trading Bot (MOEX)

Autonomous ML-powered trading system for Moscow Exchange via Tinkoff Invest API

  • ML: Ensemble of LightGBM + ExtraTrees with Optuna hyperparameter optimization; 52 technical features with per-ticker feature selection; nightly retraining
  • Finance logic: Stop-loss with dividend protection, FIFO position accounting, P&L calculation including commissions and taxes
  • Stack: Python · LightGBM · scikit-learn · Optuna · TA-Lib · aiogram 3 · SQLAlchemy async · PostgreSQL · Redis · gRPC · Docker

📅 Zapiska — Commercial SaaS: Telegram Booking Bot

Production booking service for appointment-based businesses (masters & clients)

  • Features: Service configuration, schedule management, client booking, subscription system, Tinkoff payment integration
  • Architecture: Repository pattern, FSM (aiogram States), fully async, clean layer separation, Redis caching, APScheduler for deferred tasks
  • Stack: Python · FastAPI · aiogram 3 · PostgreSQL · SQLAlchemy async · AsyncPG · Redis · APScheduler · Docker Compose · Alembic

🤖 wb-parser — AI Review Auto-Responder for Wildberries

Automated service that generates and posts AI-written replies to customer reviews

  • AI: Dual LLM support — Yandex GPT and GigaChat; runs every 30 minutes via GitHub Actions (zero hosting cost)
  • Built from experience: Developed based on personal e-commerce operations (200+ SKU portfolio, ₽46M peak revenue)
  • Stack: Python · Yandex GPT API · GigaChat API · Wildberries API · GitHub Actions CI/CD · aiogram · structlog

Full portfolio of academic and applied ML projects

Domain Projects
🔵 Classification 8 projects
📊 Regression 7 projects
🧠 Deep Learning 3 projects
📝 NLP 3 projects
👁️ Computer Vision 1 project
📈 Time Series 1 project
⚡ Big Data (Apache Spark) 1 project
🗄️ SQL 2 projects

Algorithms & Methods: Gradient Boosting (XGBoost / LightGBM / CatBoost) · Random Forest · CNN · BERT / Transformers · Word2Vec / TF-IDF · ResNet50 · Time Series (ARIMA, seasonal decomposition) · SHAP explainability · Optuna HPO


💼 Professional Background

Period Role Company
Nov 2024 – Present Data Science Specialist Freelance / Project work
Aug 2022 – Present Analyst / E-commerce Entrepreneur Wildberries Marketplace
Nov 2012 – Jul 2022 Design Engineer OKBM Afrikantov (Nuclear / Defense)

Engineering background in designing PLC-based control systems for nuclear, oil & gas, and defense industries — brings systematic thinking and attention to reliability into every ML solution.


🎓 Education & Certifications

  • 🏅 Advanced Data Science Specialist — Yandex Practicum, 2026
  • 🎓 B.Sc. Electrical Engineering — Nizhny Novgorod State Technical University (NNSTU), 2012
    Faculty of Automation & Electromechanics

📬 Let's Connect

I'm actively looking for opportunities as a Machine Learning Engineer or Data Scientist.
Open to remote positions. Ready for technical interviews and test assignments.

Telegram


💡 "Fast to learn new domains. Passionate about solving ambitious problems."

Pinned Loading

  1. yandex_practicum_projects yandex_practicum_projects Public

    This section contains 19 educational projects and 2 projects with real data that were completed during the "Data Science Specialist" course at Yandex Practicum.

    Jupyter Notebook

  2. wb-parser wb-parser Public

    Автоматический бот для ответов на отзывы Wildberries с использованием AI и Telegram уведомлениями.

    Python

  3. we_trust_in_people we_trust_in_people Public

    Автоматизированный торговый помощник для Tinkoff Инвестиции с ML-сигналами покупки/продажи.

    Python