Help us turn frontier AI into real products

What we're hiring for

We deliberately don't post boilerplate job descriptions. Instead, we look for people whose background fits one of the directions below — scope is shaped by the person and the problem, together. If any of these resonate, reach out directly.

• Generative Images / Diffusion Models

  Priority

  Model research and productionization around portrait retouching, style transfer, and controllable multimodal generation. Deep expertise in at least one of: diffusion training, inference optimization, or large-scale data curation.

  What we look for

  • ▸ Hands-on with DiT / Stable Diffusion / Flow Matching architectures
  • ▸ Can reproduce a paper end-to-end into a working PoC
  • ▸ Bonus: large-scale dataset construction or curation experience
  • ▸ Thoughtful about the cost/quality tradeoffs in inference

• AI Inference & Edge Deployment

  Priority

  Push large language and diffusion models to edge devices — quantization, distillation, sparsity, KV-cache optimization, and writing CUDA / Triton kernels are everyday work here.

  What we look for

  • ▸ Familiar with at least one inference framework (TensorRT / vLLM / SGLang / MLC-LLM, etc.)
  • ▸ Experience with INT4 / FP8 or quantization-aware training
  • ▸ Bonus: can read and modify CUDA / Triton kernels
  • ▸ Cares about actual latency and memory footprint, not paper-only numbers

• Embodied Intelligence / Multimodal Robotics

  Rolling

  Exploring VLA (vision-language-action) models, teleoperation data pipelines, and bringing world models onto real hardware. Mostly pre-research right now — looking for people willing to wade into open problems.

  What we look for

  • ▸ Background in embodied AI, robot learning, or reinforcement learning
  • ▸ Strong hands-on instincts — comfortable with real hardware or simulators
  • ▸ Clear sense of the "paper → demo → product" path

• Product Engineering / AI Applications

  Rolling

  Turn model capabilities into products users can touch — Web, mobile, SDK form factors all on the table. Backend, frontend, and product engineers all welcome; the common bar is owning a feature end-to-end.

  What we look for

  • ▸ Strong fluency in at least one mainstream language (TypeScript / Go / Python / Rust, etc.)
  • ▸ Shipped a user-facing product before — and stayed around for post-launch maintenance
  • ▸ Happy to sit next to model researchers to chase end-to-end issues, no finger-pointing

• Research Intern

  Intern

  For current Master's / Ph.D. students, mentored directly by our Chief Scientist and core researchers. Minimum 3 days/week on-site, 3+ months continuous. Strong interns can lead to a co-authored paper and a return offer.

  What we look for

  • ▸ Master's or Ph.D. in CS, AI, applied math, or related
  • ▸ Bonus: first / co-first author at a top venue or on arXiv
  • ▸ Can validate ideas with code — not just survey papers