Deep Learning Engineer (CV, LLM & VLM)
| Master’s Student @ ITMO AI Talent Hub
| 🎓 ex-Researcher @ Russian Academy of Sciences
📍 Vladivostok, Russia (GMT+10)
I build end-to-end ML systems — from data collection to production microservices — with a focus on real-time computer vision and multimodal document understanding using Vision-Language Models (VLMs).
- 🔍 Computer Vision: Object detection, segmentation, tracking (YOLO, SORT, Ultralytics) for robotics & industrial automation
- 📄 Multimodal AI: Building document processing pipelines with Qwen2.5-VL, vLLM, and Arize Phoenix
- 🏗️ Full ML Lifecycle: Data annotation → Training → Serving → Monitoring → Human-in-the-loop feedback
- 🌐 Open Source: Lead developer of VLMHyperBench — benchmark for VLMs on Russian documents
- 🎓 Academic Roots: 8+ years in scientific research (acoustics, signal processing) — published in peer-reviewed journals (Scopus, Web of Science, eLIBRARY)
| Category | Tools & Frameworks |
|---|---|
| Languages | Python |
| DL / ML | PyTorch, Lightning, Hugging Face, vLLM, Unsloth, LightAutoML, CatBoost |
| Multimodal | Qwen-VL, VLMEvalKit, LangChain, FAISS, Arize Phoenix |
| MLOps | MLflow, Weights & Biases, TensorBoard, Prefect, ONNX, TensorRT |
| DevOps | Docker (+NVIDIA), Git, uv, poetry, Docker Compose, Git SubmVodules |
| Data & CV | OpenCV, PIL, kornia, FiftyOne, CVAT, Label Studio, MinIO, RabbitMQ |
Only 3 projects are presented below, the rest can be found at the link bellow:
Multimodal system for real-time emotion detection in Zoom/Skype calls.
🥇 1st place in ODS MLOps course • Demo video with live inference
Open benchmark for evaluating Vision-Language Models on Russian documents.
🏆 Winner of Yandex Open Source Grant 2025 • Presented at Data Fest 2025
Production service for extracting structured data from Russian documents using multimodal LLMs.
🔁 Human-in-the-loop feedback • Built with vLLM, MinIO, RabbitMQ, Arize Phoenix
- Master’s in AI, ITMO University (AI Talent Hub) — courses in Multimodal Models, ML System Design, Model Compression
- PhD in Physics & Mathematics (Acoustics, RAS) — 8+ years in research, Scopus/WoS publications
- 🏆 Yandex Open Source Grant 2025
- 🥇 1st place, ODS MLOps Track
- 🥉 Top-3, AI Talent Hackathon 2023
- 🎓 Selectel Career Wave Scholarship (2023, 2024)
📧 Email: [email protected]
📄 Full CV: Google Drive
💡 "All the most interesting things happen at the intersection of fields."
— From acoustics to multimodal AI.