AI Engineer | LLM Agents, RAG, Knowledge Systems & Model Inference
I design and build AI systems from prototype to production, with a focus on LLM agents, multi-agent orchestration, intelligent search, retrieval evaluation, and efficient model serving.
I currently work as an AI Engineer in R&D and pursue a master's degree in Multi-Agent Intelligent Systems.
- LLM agents and multi-agent workflows
- RAG, hybrid retrieval, reranking, and enterprise knowledge systems
- LLM evaluation, model serving, and inference optimization
- Backend services and deployment infrastructure for AI applications
My bachelor's thesis project: a LangGraph-orchestrated system of specialized agents for U.S. equity research and portfolio decision support. It combines market, fundamental, technical, and news analysis with RAG-based decision memory, portfolio construction, risk controls, explainability, and execution tracing.
- Agentic systems: LangGraph, OpenHands, Haystack, multi-agent orchestration, tool-driven workflows, structured outputs with Pydantic
- RAG and knowledge systems: hybrid retrieval (BM25 + dense search), BGE-M3 embeddings, BGE-Reranker-v2-M3, ChromaDB, TF-IDF, document ingestion pipelines, retrieval evaluation
- Model serving and ML: vLLM, PyTorch, Hugging Face Transformers, XGBoost, scikit-learn, inference benchmarking (TTFT, TPOT, throughput, concurrency, KV-cache utilization)
- Backend and data: Python, FastAPI, REST APIs, SQL, PostgreSQL, Redis, MinIO, Node.js
- Infrastructure and delivery: Docker, Linux, GitHub Actions, GitLab CI/CD, Streamlit
- M.S. in Information Systems and Technologies, Multi-Agent Intelligent Systems — RTU MIREA, expected June 2028
- B.S. in Fundamental Informatics and Information Technology, Artificial Intelligence and Machine Learning — RTU MIREA, June 2026