Design and train foundation vision-language models that read entire CT studies end-to-end. Build custom PyTorch architectures and distributed training pipelines, run large-scale experiments, evaluate models for clinical safety, and move models through FDA submission to hospital deployment.