I need a full-stack developer who can build an AI-driven application that lets me do two things in one place: first, generate complete music videos, and second, train and manage my own AI models. For the music-video side, the user should be able to pick a visual style and drop in a custom voice track that the system blends seamlessly with the chosen look. Adding extra video effects is something I may explore later, so please keep the architecture flexible enough to plug that in without a rewrite. On the modelling side, I want a workflow that lets me start a new model from scratch, feed in data, train, save, and then reuse that model inside the video generator. Ideally this is handled behind a clean dashboard rather than command-line steps. Core expectations • Intuitive UI/UX for both the video creator and the model-training console • Scalable backend (Python, PyTorch/TensorFlow, or comparable) with GPU support • Real-time preview for voice and style selections • Secure user accounts and storage for datasets, models, and rendered videos • Source code, build instructions, and a short hand-off call on completion If you have a reusable codebase or experience with generative media tools such as Stable Diffusion, FFmpeg, or Jukebox-style audio models, let me know—speed of development matters.