Tech Stack: Python, C++, PyTorch, Tensorflow, LLMs, NLP, Computer Vision, Git, Linux, TFLite, Visual Transformers, Reinforcement Learning, Object Detection, Object Tracking, Semantic Segmentation, Depth Estimation, SLAM, GANs, Time Series Analysis, SQL, MapReduce, Machine Learning, Neural Networks, XGBoost, CatBoost, Robotics, Big Data, Scikit-learn, CI/CD, Unit Tests, Kaggle, MLFlow, DVC, Kubernetes, Docker

Expert’s cases:
1. Enhancing Safety Driving and Drivers Scoring
- Installed dual cameras in vehicles for simultaneous analysis of road and driver behavior
- Developed a system for evaluating driver performance and identifying dangerous driving
- Main result: Identified dangerous driving with 82% accuracy

2. In-house LLM Knowledge Base Localization
Created an in-house ChatGPT analogue for a proprietary knowledge base
Developed a cloud-ready solution for data ingestion, training, and inference

3. Next-gen learning
Provide the faculty tools and techniques to enhancе student experience and academic performance, including engagement, bullying
Main results:
- Students engagement rate 74%
- Student performance prediction to the end of the month - precision 95%
- Bullying detection with 90% accuracy
- Topics clustering with 95% classification accuracy

4. Plasma Control in Nuclear Fusion Using Reinforcement Learning
The project aimed at optimizing the control algorithms for stable plasma confinement, crucial for sustainable nuclear fusion processes
Main results:
- Enhanced Plasma Stability
- Cost Reduction
- Scalability