Rustam Uzdenov SUPERustam

Hi, I’m Rustam Uzdenov 👋

🧠 Junior Computer Vision Researcher at AIRI and Moscow State University (MSU)
🎓 Bachelor's degree, BMSTU '26 (Applied Math/CS) - GPA 4.35/5
🔬 Focus: Deep Learning, Computer Vision, GenAI for CAD, 3D Geometry, Reverse Engineering, LLM

Junior Computer Vision Researcher @ AIRI (Sep 2025 - Present)
- Computer Vision Research: Generative AI for CAD, 3D Computer Vision
- Reverse engineering: parametric reconstruction of CAD models from meshes or scans
- 3D Computer Vision: surface reconstruction, correspondence estimation, point cloud/mesh registration, feature extraction.

Joint AIRI, MSU, Sber research project.

Junior Computer Vision Researcher @ Moscow State University (MSU) (Sep 2025 - Present)
- Computer Vision Research: Generative AI for CAD, 3D Computer Vision
- Reverse engineering: parametric reconstruction of CAD models from meshes or scans
- 3D Computer Vision: surface reconstruction, correspondence estimation, point cloud/mesh registration, feature extraction.

Joint AIRI, MSU, Sber research project.

Junior Computer Vision Researcher @ ISP RAS (Sep 2024 - Sep 2025)
- R&D in Computer Vision with a focus on Video Copy Detection.
- Published and presented research on "Shazam Algorithm for Partial Video Copy Detection" at the IVMEM2025 conference.
Machine Learning Engineer @ Fast Food Memes (Jan 2024 - Aug 2024)
- Developed an infinite personalized meme feed for an open-source Telegram bot (repo).
- Processed 500k+ memes with OCR in 10 languages, reaching 2,500 MAU.
- Built a deduplication and spam-detection system that flagged over 14k spam entries.
MLOps Engineer @ Bauman Racing Team (Nov 2022 - Oct 2023)
- Created a CLI tool to manage Docker containers for a driverless car system.
- Developed a C++ unit testing system and built CI/CD pipelines from scratch.

Programming: Python, SQL, C++, Bash
ML/DL: PyTorch, Transformers 🤗, OpenCV, Open3D, PEFT, vLLM, NumPy, Pandas, Matplotlib, Scikit-learn etc.
Technologies: LoRA, Qwen, GraphRAG, RAG and more!
Collaborative Skills: Git & GitHub, Docker, Airflow/Prefect, Leadership, Research Collaboration, Code Review
Languages: English (Professional Working), Russian (Native)

Optimizing GraphRAG for Enhanced RAG Pipelines (July 2025)
- Team project at StudCamp on NLP Yandex × NSU.
- Optimized GraphRAG construction using LoRA fine-tuning on Qwen3-4B for enhanced NER and relation extraction in Russian.
- Implemented the solution into the open-source RAGU framework.

Yandex StudCamp on CV & Autonomous Transport with Faculty of CS HSE — Yandex, Oct 2025 Credential
Yandex StudCamp on NLP with Novosibirsk State University (NSU) — Yandex, Jul 2025 Credential

Always learning, always building.