Introducing Moonlake, AI for creating world simulations.

Modeling & architecture

• Build and iterate on 2D/3D/image/video/audio diffusion architectures

• Work on conditioning: text/image/pose/layout/control signals, multi-modal encoders, guidance strategies.

Training & optimization

• Large-scale diffusion training

• Improve sample quality vs. compute trade-offs: better objectives, distillation, consistency/one-step models.

Control & alignment

• Techniques like ControlNet, LoRA, IP-Adapters, adapters for style/identity/geometry/control.

• Build editing, inpainting, and personalization pipelines (DreamBooth, custom subject/style tuning).

We are committed to being an on-site, in-person team currently based in San Mateo