Vladyslav

VLADYSLAV ZELINSKYI
[відкрити контакти](див. вище в блоці «контактна інформація») • https://github.com/Wertyhan
• [відкрити контакти](див. вище в блоці «контактна інформація») • Rivne, Ukraine

SUMMARY
AI Engineering Student (4th year) with 1+ year of practical experience in building end-to-end RAG systems,
MLOps pipelines, and Agentic RAG workflows.

EXPERIENCE & PROJECTS

Stratt(AI Engineer / Backend Developer | Duration: 2 months | Team size: 2 developers)
• Designed and implemented a multi-stage AI pipeline (slide -> markdown -> insights) with strict quality
gates and schema validation.
• Built robust taxonomy control and enum normalization to keep AI outputs consistent and enterprise-
safe.
• Optimized high-volume document ingestion by implementing async processing, batching, rate-limit
control, and concurrency tuning.
• Developed API endpoints and background extraction flows.
• Automated end-to-end data delivery by seamlessly integrating external systems, including S3/MinIO
and Airtable.
• Implemented traceability and observability with structured logs, token usage tracking, and processing
diagnostics.
• Prepared the platform for retrieval and chatbot use cases by producing clean, metadata-rich, RAG-
compatible insight records.

BeliefCraft(AI Engineer / Backend Developer | Duration: 4 months | Team size: 8 developers)
• Architected a multi-service AI platform that seamlessly integrates agent orchestration, semantic
retrieval, and warehouse domain APIs.
• Built a ReAct-based decision workflow using LangGraph and AWS Bedrock to support iterative
reasoning, dynamic tool calling, and traceable answers.
• Developed MCP-based RAG integration, enabling dynamic tool discovery and semantic knowledge
base retrieval with graph expansion.
• Implemented robust FastAPI microservices along with shared backend packages for typed contracts,
request tracing, and inter-service communication.
• Designed warehouse data access flows using PostgreSQL and SQLAlchemy, including smart-query
endpoints for inventory and device monitoring.
• Ensured system reliability and observability through structured logging, LangSmith-compatible LLM
tracing, and comprehensive testing (pytest, Testcontainers).

SKILLS
Core & Backend: Python 3 (AsyncIO, Pydantic, NumPy, Pandas, Scipy), C# (Basic), C++ (Basic), REST
API (FastAPI, Uvicorn), Docker, Git.
AI & LLM Stack: LangChain, LangGraph, AutoGen, RAG, Agentic RAG, Embeddings, Semantic Search,
Prompt Engineering.
Databases: SQL, PostgreSQL, Vector Databases (ChromaDB, pgvector concepts).
Languages: English (Upper-Intermediate / B2), Ukrainian (Native).
Cloud Platforms: Azure (Azure OpenAI, Azure Bot Service, App Service, Blob Storage), AWS (Bedrock,
Lambda, S3), Google Cloud Vision API

EDUCATION
Kharkiv National University of Radio Electronics Bachelor's degree in Artificial Intelligence (4th Year)