Oleh Semenets

Designed and optimized LLM prompts and evaluation workflows to improve reliability of AI outputs. Built retrieval-augmented generation (RAG) pipelines using embeddings, sparse/dense retrieval, and caching strategies. Developed agentic workflows with LangChain and LangGraph, improving automation and tutoring system personalization. Integrated Node.js and Python Flask APIs for scalable backend services. Enhanced application resilience with robust error handling, observability, and monitoring pipelines.

Delivered AI-driven features using LLMs and agent-based architectures, enabling workflow automation and personalization. Implemented Model Context Protocol (MCP) for secure multi-system integration. Built AI orchestration pipelines with n8n and Make to automate complex workflows. Collaborated with stakeholders to design AI-powered product features aligned with B2B and B2C business goals. Deployed scalable cloud-native AI systems across AWS and GCP, ensuring high availability and cost efficiency.

Led backend team building deep learning–driven trading bots for real-time financial decision-making. Engineered data pipelines for preprocessing, feature engineering, and model validation. Developed Flask-based microservices on AWS/GCP and integrated SQL/NoSQL data stores. Built analytics dashboards and Tableau reports to evaluate and monitor ML model performance.