Open to opportunities

Jevgēnij Posaškov

@jevgnijposakov

Message

I build interpretable machine learning systems for vision, probabilistic modeling, and edge deployment.

Slovenia

Message

What I'm looking for

I'm looking to build efficient, interpretable machine learning systems, especially in spatial AI, depth-aware computer vision, model optimization, and edge-device deployment.

I've developed interpretable ML systems with Siemens and the University of Ljubljana for EV charging using real-world grid data, deep learning, SHAP, and Logic Tensor Networks. The work achieved over 90% accuracy, identified key influencing factors, and reached finalist status in an industrial AI competition.

At Orange Data Mining, I built a Bayesian inference framework in Stan and Python for automated MCMC and hierarchical modeling, alongside validation pipelines for posterior checks and sensitivity analysis. I also designed GUI tools that help non-expert users apply probabilistic models while improving interpretability and reproducibility.

My projects span RAG with Mistral 7B, FAISS, and Wikidata; neuro-symbolic reinforcement learning; and computer vision models including ResNet-18, U-Net, and FCN. I'm particularly interested in efficient neural networks, spatial AI, depth-aware vision, model optimization, and edge ML deployment.

Experience

Work history, roles, and key accomplishments

ML Researcher (Bayesian Statistics)

Orange Data Mining

Dec 2024 - Dec 2025 (1 year)

Built a Bayesian inference framework using Stan and Python for automated MCMC and hierarchical modeling. Developed a model validation pipeline including posterior checks and sensitivity analysis.

Stan Python Hierarchical Modeling Model Validation GUI Design

Data Scientist

Siemens

Feb 2024 - Jun 2024 (4 months)

Developed interpretable ML models for an EV charging system using real-world grid data. Combined deep learning, SHAP analysis, and Logic Tensor Networks (LTN) to extract decision rules and operational thresholds.

Deep Learning SHAP Logic Tensor Networks Python Machine Learning Interpretation EV Charging Data Analysis

Data & Sales Analyst

Prospect CRM

Feb 2023 - Oct 2023 (8 months)

Built Python data pipelines for large-scale lead extraction. Performed funnel optimization using SQL and improved conversion analytics and reporting automation.

Python SQL Data Pipelines Funnel Optimization Conversion Analytics Report Automation

Research Assistant

Bauman Moscow State Technical University

Sep 2021 - Aug 2022 (11 months)

Modeled nano-optical components using Python and FDTD simulations. Optimized photonic circuit performance and designed architectures for quantum computing applications.

Python Quantum Computing