Build, scale, and 'reliabilize' an infrastructure to support a cluster of >1000 GPU-nodes performantly, simply, and beautifully. Requires excellent understanding of distributed systems and engineering trade-offs.
A company of 6 engineers making a new type of creative tool. Older creative tools wrapped computer graphics advances into a performant controllable user interface (e.g. brightness, hue, saturation, etc.) and we think modern tools should do the same for the latest AI research (e.g. diffusion models).
