Arseniy Glotov

• Building and deploying RAG pipeline with LLM• Building MLOps architecture for a ML team

Working on projects for FivePost, a logistics company with 20 million parcels delivered in 2022. • Built ML system at scale for Voice Assistant Bot to reach the goal of 50%+ of customer problems solving without the need for operator assistance; • Implemented a streamlined model training pipeline with integrated MLOps and CI/CD tools, cut- ting the development time for new classifier models from 6 weeks to 2 weeks; • Refactored and improved the ML system with classification & NER models;

Three-month contract position. • Deployed a keyword extraction model into a production ML service, improving the company’s search engine product; • Built asynchronous web parsers for several websites, resulting in significant data enrichment and quality improvements in the search engine product.

Worked on a project to build a cashierless store for the Russian retail company ”VkusVill”. • Deployed object detection models on embedded devices; • Developed a Person Re-Identification system to improve customer tracking algorithms by lowering tracking errors by 15%.

Project-based work for local companies. • Applied forecasting algorithms to predict bookmaker’s profits and detect unprofitable offices; • Collected and analyzed social indicators in different regions of Russia, using clusterization algo- rithms to identify clusters of similar regions; • Developed and deployed a REST API service for working with geographic data.

Thesis: «Pulmonary fibrosis progression prognosis using machine learning»

Grade: summa cum laude