Fedor Pakhurov
AI / ML Engineer focused on LLM systems, agents, retrieval and applied AI products.
Summary
AI / ML engineer building systems around large language models, retrieval, agents and multimodal models, with an emphasis on production integration and operational usefulness.
Selected experience
Applied LLM systems
Production-oriented NLP and LLM systems for e-commerce operations, spanning RAG, agents, multimodal prototypes and internal AI services.
Multi-agent inventory operations
Designed a coordinator-led platform with specialist supply, transit and export agents, tool access and explicit confirmation for critical operations.
AI transformation
Worked on AI-first maturity strategy linking process digitization, decision data, business rules, ML platforms, LLM systems and agents.
Research & education
Methods of Evaluation of Diffusion Model Hallucinations
Bachelor thesis covering diffusion models, complexity–entropy analysis, Rényi entropy, Bandt–Pompe methods, clustering and hallucination severity estimation.
Technical skills
Applied AI
LLM systems, RAG, agents, multimodal models, evaluation
ML engineering
NLP, generative models, experimentation, MLOps, production integration
Systems
Architecture, tool interfaces, human-in-the-loop workflows, operational AI