Looking for a job

ALEKSEY Deev

@alekseydeev

Data Scientist specializing in marketplace analytics, production ML systems, and NLP/LLM solutions. Open to remote roles.

Russian Federation

Message

What I'm looking for

I’m looking for a remote Data Science role where I can build leak-free, production-grade ML for marketplace problems, including fraud detection and LLM-powered NLP features, with freedom to iterate quickly and measure impact.

I am a Data Scientist and ML Engineer based in Nizhny Novgorod, Russia, with a strong engineering background and hands-on experience building production AI systems.

My technical journey combines nearly 10 years as a control systems engineer at OKBM Afrikantov (Rosatom): designing PLC-based systems for nuclear, oil & gas, and defense industries with a deep transition into Data Science and Machine Learning. I completed an Advanced Data Science program at Yandex Practicum (2026) and have built several production-grade projects from scratch.

Key projects include an algorithmic trading bot for the Moscow Exchange featuring a custom RankEnsemble (LightGBM + ExtraTrees + HistGradientBoosting) with Walk-Forward validation, an AI auto-reply service for Wildberries marketplace reviews using YandexGPT and GigaChat APIs, a RAG-powered AI chatbot with pgvector semantic search and per-user memory, and a SaaS Telegram bot for client scheduling with payment integration.

I also ran my own e-commerce business on Wildberries and Ozon for three years, reaching ₽46M peak annual revenue with 200+ SKUs: giving me deep understanding of marketplace dynamics, fraud patterns, pricing elasticity, and data-driven decision making.

My primary development tool is Claude Code, which I use daily. I am comfortable working autonomously, shipping production-ready solutions, and explaining technical results to non-technical stakeholders.

I am open to remote Data Scientist, ML Engineer, and AI Engineer roles.

Experience

Work history, roles, and key accomplishments

Current

Data Science Specialist

Current

Project-Based Work

Nov 2024 - Present (1 year 7 months)

Built production ML/NLP systems and marketplace fraud/market data models using Python, LightGBM/XGBoost/CatBoost, and Transformers, including leak-free walk-forward validation (gap=4, embargo=4) and SHAP-based interpretation. Delivered applications integrating LLMs (YandexGPT/GigaChat) for MOEX trading and Wildberries review automation with scheduled execution and daily performance reporting.

Python XGBoost Optuna HPO BERT Transformers (Hugging Face)Walk Forward Validation SHAP

Current

Marketplace Sales Analyst

Current

Independent Practice

Aug 2022 - Present (3 years 10 months)

Managed Wildberries marketplace operations and performed sales, demand, and inventory analytics while setting up and optimizing advertising campaigns. Achieved peak revenue of ₽46M in 2023 and grew to a portfolio of 200+ SKUs, with automated reporting via Google Sheets API.

Inventory Analytics Marketing Optimization Wildberries Seller Operations performance reporting