Work Experience

Continuing development of schema-driven code generation system for the DPF simulation framework. The system follows traditional compiler architecture with front-end parsing, intermediate representation, and back-end code emission targeting FlatBuffers format.

Master's thesis internship: designed and implemented a multi-language code generation pipeline for the DPF simulation framework, replacing manually maintained client libraries in Python, C++, and C# with auto-generated code from schema definitions.

Developed mutation testing tools for a large-scale C++ codebase (ticket search engine), involving AST construction, partial compilation, and code mutation to assess test-suite effectiveness.

Compared neural network (DROW) and algorithmic person detectors for 2D LiDAR data on the RobAIR platform. Developed the FTD comparison library, ROS package, and Docker deployment image.

Software engineer on the Chatsky open-source conversational AI framework. Developed neural network model wrappers, CI/CD pipelines, and documentation. The framework relies heavily on asynchronous and multithreaded architecture for efficient distributed LLM request processing.

Graduated with 15.343 of 20 average score (French grading system). Notable courses: Programming Languages and Compiler Design (14.1), Principles of Operating Systems (17.45), Introduction to Distributed Systems (16.75), Computer Networks Principles (17.24), Parallel Algorithms and Programming (15.6).
Exchange semester at HKUST (16.0/20): Computer Security, Computer Networks, Deep Learning for Computer Vision, Distributed AI Systems.
Master's thesis: researched binary schema specification formats (FlatBuffers, Protobuf, Cap'n Proto, MessagePack) and designed a schema-driven code generation pipeline for multi-language client libraries in the Ansys DPF framework; supervised by Camille Bellot.

Graduated with 4.4 of 5.0 average score (Russian grading system). Specialization: Development of Software Information Systems. Notable courses: Operating Systems (A), Networks and Telecommunications (A), Algorithms and Data Structures (A), Mathematical Logic and Theory of Algorithms (A), Cryptography and Information Protection (A), Parallel Algorithms (A), Machine Learning (A), Assembly Programming (A).
Bachelor's thesis: designed and built a crowd-sourced monitoring system for Sosnowsky's hogweed, an invasive phototoxic plant species. The system combines a Django/PostGIS backend, a Flutter mobile client for field reporting, and a TensorFlow Lite image classifier for automated species identification; grade: excellent.