Андрій

custom racket characteristics (speed, spin,
ANDRII BOLONNYI control) based on blade and rubber materials.
●​ Integrated Google Gemini 2.5 Flash for
generating context-aware expert responses and

📍
AI Engineer / Data Scientist / Python Backend utilized MongoDB for user state and custom
Developer Lviv, Ukraine (Remote / Hybrid) assembly management.

📧 [відкрити контакти](див. вище в блоці «контактна інформація») ●​ Published the pre-processed text corpus on
Kaggle with a high Usability Score, providing a

🔗 [відкрити контакти](див. вище в блоці «контактна інформація») starter notebook for community exploratory data
analysis (EDA).
https://www.kaggle.com/datasets/andriybolonnyi
NoSQL Music Analytics System | MongoDB,
https://github.com/Andre7425 High-Volume Data, Aggregation
●​ Designed a high-performance schema for artist
PROFESSIONAL SUMMARY & trend analytics using complex Aggregation
Pipelines.
Analytical, data-driven, and business-minded ●​ Optimized queries through compound indexing,
professional with a solid foundation in Applied reducing response time for heavy analytic tasks.
Mathematics, currently pursuing an MSc in Computer
Science (AI & ML track). Proven track record of Taxi Ride Data Pipeline | Python, MongoDB, Git
architecting scalable backend systems, specialized
automation bots, and data engineering pipelines. ●​ Built an automated pipeline for cleaning and
Expertise in data analytics, vector similarity metrics, processing large-scale mobility datasets.
mathematical modeling, and NoSQL aggregation ●​ Implemented spatial-temporal filtering to extract
frameworks. Combines strong technical capability with operational business insights.
hands-on entrepreneurial acumen to bridge the gap
between complex algorithmic structures and practical, Numerical Programming | Python, NumPy, Linear
production-grade AI solutions for diverse business Algebra
needs. ●​ Programmed matrix decompositions and
optimization algorithms from scratch.
CORE EXPERTISE & TECHNICAL STEK ●​ Built the mathematical foundation for custom
linear regression and ML models.
●​ Programming & Backend: Python (OOP,
Asyncio), SQL, Shell Scripting, Flask, REST Hybrid Semantic Search Engine (RAG Pipeline)
APIs, Webhooks.
●​ Data Science & AI: Machine Learning ●​ Tech Stack: Python, Pinecone, HuggingFace
Fundamentals, RAG Systems, Vector Search, (specter2), rank-bm25, Pandas.
Linear Regression, Predictive Modeling, ●​ Engineered an end-to-end search pipeline for a
Distance & Similarity Metrics, Geometric Data dataset of 10,000+ scientific papers, optimizing
Analysis. data ingestion via semantic chunking and
●​ Databases & Infrastructure: MongoDB Parquet storage.
(Advanced Aggregation Pipelines, Schema ●​ Implemented a scalable vector search
Design), Pinecone, Qdrant, Vector DBs, Docker architecture using a domain-specific embedding
Desktop, Git, Linux. model and Pinecone for fast nearest-neighbor
●​ Business & Soft Skills: Product Mindset, retrieval and metadata filtering.
E-commerce Operations, Analytical Problem ●​ Maximized search relevance by combining
Solving, Data-Driven Decision Making. semantic embeddings with sparse lexical search
●​ Languages: English (A2+), Ukrainian (Native). (BM25), merging results via the Reciprocal Rank
Fusion (RRF) algorithm.
KEY PROJECTS / PORTFOLIO
PROFESSIONAL EXPERIENCE
AI Table Tennis Coach & Racket Customizer Bot |
Python, LangChain, Pinecone, Gemini API, Regional Manager | Commercial Sales &
Scikit-learn, MongoDB Operations Sector | 2010 — 2023

●​ Developed and deployed a full-stack Telegram ●​ Coordinated logistics, local distribution
bot combining a domain-specific AI coaching networks, and supply chains across two regional
assistant (RAG) and a machine learning-driven commercial entities.
racket customizer. ●​ Designed internal reports and data tracking
●​ Built a custom data pipeline using LangChain to workflows to predict quarterly market demands
clean, semantically chunk, and vectorize and streamline operational budget management.
unstructured sports literature into Pinecone
Vector DB for precise semantic search.
●​ Trained and serialized a predictive regression
model (Scikit-learn, Joblib) to instantly forecast
EDUCATION & CREDENTIALS

●​ MSc Computer Science (AI) Woolf
University | In Progress Focus: NoSQL
and Vector Databases, ML Fundamentals,
Scalable AI Infrastructure, Advanced Data
Models.
●​ MSc Applied Mathematics I.Fanko LNU
of Lviv | Conferred Focus: Mathematical
Statistics, Optimization, Numerical
Algorithms.

PERSONAL INSIGHTS / INTERESTS

●​ Data Driven, Sports Analytics: Applied vector
math models to table tennis equipment
behaviors, creating a real-world AI
recommendation tool.