Engineering
Computer Vision Engineer
Design and ship large-scale vision models
About the role
As a Computer Vision Engineer, you'll design, train, evaluate, and ship large-scale vision models, while also making the underlying systems blazingly efficient. You'll own projects end-to-end: data curation and augmentation, training loops, distributed pipelines, inference graphs, serving optimizations, and measurement frameworks.
What you'll do
- Train, evaluate, and ship vision models that run in production, not just in notebooks
- Own the data: curation, augmentation, and the evaluation sets that tell you the truth
- Build distributed training pipelines and keep them efficient as they scale
- Optimize inference graphs and serving so generation stays fast and affordable
- Build the measurement frameworks that decide whether a model change is actually better
You may be a good fit if
- Strong software engineering fundamentals, whatever your background
- Experience training and deploying deep learning models at scale
- Fluency with PyTorch and GPU performance work
- A bias toward measuring things rather than arguing about them
Nice to have
- Generative image models, diffusion, or multimodal work
- Large-scale ETL and data pipelines
- Inference optimization: quantization, batching, kernel-level work
About NEX
NEX is the growth system for consumer packaged goods brands. We replace the scattered stack of agencies and point tools with one place where creative, measurement, and media buying feed each other: ad performance tells us what to make next, and what we make gets measured against net profit rather than platform-reported ROAS. We're a small team working directly with the brands that use us.
Apply even if you don't match every line above. The lists describe the work, not a checklist. If you can do the job, tell us how you know.
Apply for Computer Vision Engineer
Attach a resume or share a link. Everything goes straight to the team that's hiring.