We generate in ~2 seconds. We want faster. Speed isn't a nice-to-have — it's what makes creation feel like play. You'll own the full ML pipeline and work directly with Josh, who's deep in infra and will be right there with you. No research-only work here — you own production. What you'll do- Own inference end-to-end — speed, quality, cost- Run and optimize diffusion models in production- Build MLOps — rollouts, A/B testing, fallbacks, monitoring- Evaluate new models fast and ship the ones that win Tech stack- Python / PyTorch- Hyper-fast diffusion models and low-latency LLMs- AWS GPU infrastructure, inference optimization What we want- Cares about what users feel, not just benchmark scores- Owns production, not just experiments- Thinks in systems: latency, cost, quality, scale- Takes ownership by default