🌐 Live portfolio: vados182.github.io/mkdocs-portfolio
This repository contains the MkDocs Material source for my portfolio. It presents my work as an Industrial AI Quality Specialist, including a work prototype for quality-document automation, independent data/AI prototypes, and analytical projects.
Project pages distinguish prototypes and demonstrations from production deployments and describe known limitations. The FPS page is a sanitized overview: no supplier documents, employee records, credentials, or internal configuration are published.
- FPS Quality Document Automation: Work prototype for AI-assisted supplier document parsing, data validation, traceability, and employee QA review. JSON export is an integration starting point; there is no direct IFS Cloud synchronization.
- E-Commerce Customer Risk ML Pipeline: Olist analytics and XGBoost workflow. The model target is low review score, not observed customer churn.
- Multimodal E-Commerce Intelligence: Demonstration prototype combining PySpark ETL, an LLM product-analysis agent, image classification, and a Streamlit UI; sample ETL data is synthetic.
- Olist EDA and RFM Segmentation: SQL-based e-commerce analysis and customer segmentation.
- TikTok Text Sentiment Dashboard: Scheduled analysis of text already stored in PostgreSQL; it does not collect TikTok posts.
- Browse the full portfolio.
Python, SQL, Pandas, PySpark, DuckDB, PostgreSQL, SQLite, Parquet, scikit-learn, XGBoost, PyTorch, OpenAI and Gemini APIs, LlamaParse, FastAPI, Streamlit, Docker, MLflow, Azure Machine Learning, and GitHub Actions. Tools are listed as used in the projects; their presence does not imply production experience with every technology.
- GitHub: github.com/Vados182
- LinkedIn: linkedin.com/in/wadim-worobiow-226a58147
- Email: bassuha@gmail.com