Vladimir Vorobiev

I’m an Applied AI Engineer and technical product leader with 3+ years of experience building production LLM, RAG, search, and agentic systems in healthcare and legal technology.

At Humarin, I led AI architecture and product delivery from an award-winning prototype to registered software, clinical pilots, healthcare integrations, and the company’s first commercial contract. I built Humarin.Transcribe, which turns doctor–patient conversations into structured medical records and integrates them with EHR workflows, and Humarin.Diagnose&Treat, an explainable agentic clinical decision-support system grounded in patient data, medical knowledge, and clinical guidelines. In a Fomin Clinic pilot, Transcribe reduced appointment-documentation time by 50%.

I also led Humarin to win the national AI’m Doctor competition, securing RUB 35M in development funding.

Previously at Sberbank, I built legal-domain LLM, information-retrieval, search, and conversational AI pipelines for GigaLegal. I also created an open-source T5 paraphrasing model with 3.5M+ downloads.

My strongest areas are RAG, information retrieval, NLP, LLM applications, AI agents, evaluation, Python, APIs, product architecture, and translating complex domain workflows into practical AI systems.