Viktoriya Kuryla

Built and deployed AI/ML pipelines; advanced research in multi-agent AI, built an end-to-end ML pipeline; Founded and managed a dance school with around 120 students, overseeing operations; Organized public dance events and workshops.

Embedded and optimized real-time AI solutions on restricted hardware; developed and deployed ML models for creative and retail applications; built an end-to-end ML lifecycle pipeline and supported the foundation of traffic analysis software. Led the setup of a compact ML pipeline, including data augmentation, quality checks, and model deployment.

Built and deployed an end-to-end ML pipeline covering data augmentation, quality inspection, model training, pruning, retraining, evaluation, and quantization (PTQ/QAT). Deployed real-time object detection models (YOLO, DETR, ViT) across hundreds of devices and contributed to traffic analysis software foundations. Tech stack included Python, PyTorch, OpenCV, CUDA, ONNX, TensorRT, Docker, AWS, and NVIDIA Jetson platforms.