Aleksandr Algazinov

Aleksandr Algazinov

AI Engineer at Sberbank. 4+ years of ML experience (solving complex problems at top banks and doing research at Tsinghua University), specializing in LLM Agents, NLP, and applied AI in finance.

Hi, I’m Aleksandr!

Illustation of the life philosophy

My work sits at the intersection of applied AI engineering and research. As an AI Engineer at Sberbank, I design and implement end-to-end multi-agent systems and LLM applications for project management and product analysis automation. Prior to joining the largest bank in Russia, I spent over a year as an AI Researcher at Tsinghua University’s THUIAR lab, working on robust multimodal intent recognition and multi-agent architectures. With 4+ years of ML experience, my applied domains span LLM Agents, NLP, Classical ML, Ranking, Time Series, MLOps, and Applied AI in Finance.

My research has resulted in three publications: TRACE (WWW 2026), a transparency-focused reliability scoring framework for web content (GitHub); MATE, a multi-agent system for context-aware modality conversions (GitHub); and a Springer book chapter on AI in network devices.

Before coming to China, I studied Economics at HSE University, which is the best university in Russia for Computer Science and Economics. Already at the beginning of my studies, I became interested in Machine Learning. Hence, besides my main course work, I took many courses from the Department of Computer Science, such as Machine Learning, Deep Learning, Natural Language Processing, Big Data, etc.

In addition to working and doing my research, I enjoy sharing my knowledge and experience with others. For instance, I was a speaker at the Open Data Science conference (in Russian), where I talked about the methods we used to detect time series anomalies, the challenges we faced, and the approaches we took to overcome them. On the top of that, I run a telegram channel (in Russian), where I share my experience in terms of AI, studying, and living in China.

Please feel free to check my CV to know more about my work, research, and project experience, or to have a look at my research projects in the respective sections. If you’re interested in my experience, and would like to work together or to collaborate on the projects, please drop me a message via mail or telegram.