ML Engineer | Applied AI for industrial, scientific and engineering systems
I build practical ML systems for chemical technology, oil & gas, document intelligence, graph optimization and scientific computing.
Telegram | GitHub | Saint Petersburg
I work at the intersection of machine learning, chemical engineering, industrial data analysis and graph optimization.
My projects focus on applied ML systems: predicting physical and chemical properties, building RAG pipelines, working with satellite imagery, optimizing routes for infrastructure networks and turning research prototypes into usable tools.
I am especially interested in AI for oil & gas, chemical production, engineering decision support and industrial automation.
| Project | Area | What it does |
|---|---|---|
| NeftecodeTeamRocket | Industrial ML | Predicts Daimler Oxidation Test results for multi-component oil formulations. 1st place Neftecode 2026 project. |
| SteinerRL Pipeline | Reinforcement Learning / Graph Optimization | Grid-based Steiner Tree pipeline for pipeline network design experiments using A*, GCN policy and PPO-style training. |
| Optics-Hackathon | Scientific Computing / Optimization | Generates optical schemes with target physical constraints using genetic algorithms and ray tracing. 1st place hackathon project. |
| Gagarin Sentiment Interface | NLP / Financial Text Mining | Detects stock market issuers in Telegram/news texts and predicts sentiment with TF-IDF and classical ML. 3rd place hackathon project. |
- Winner of Neftecode 2026 with an ML solution for oil formulation analysis.
- Winner of Computational Optics and Imaging Hackathon 2023.
- Prize winner of Gagarin Hackathon 2024 for financial sentiment analysis.
- Winner of AI Champ Hackathon 2026 with a RAG system for video fragment search.
- Speaker at Congress of Young Scientists 2026 with a project on Steiner Tree optimization for pipeline networks.
- Practical experience with ML services, FastAPI, Docker, RAG pipelines, satellite image segmentation and industrial data.
Python | PyTorch | TensorFlow | pandas | NumPy | scikit-learn | FastAPI | Docker | Streamlit | MLflow | ChromaDB | GeoPandas | A* | PPO | GCN | RAG | LLM pipelines
- Satellite image segmentation with U-Net for industrial infrastructure planning.
- Route construction over terrain masks using desirability maps, A* and Steiner Tree ideas.
- RAG systems based on Llama, bge-m3 embeddings, ChromaDB and retrieval pipelines.
- ML models for chemical and oil & gas tasks, including PVT modeling and process optimization.
- Deployment of ML services with FastAPI and Docker.
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.