Senior Research Engineer — Applied Vision Models

Cortex AI

Remote (DE) / Hybrid (Munich, DE) · $150K – $178K · Full-time · Posted 6 months ago

This role is no longer accepting applications

Don't miss the next role like this

Create a free profile with your skills. We match you to new AI roles as they open, and companies hiring can contact you directly.

Create free profile, 2 minutes

Similar open roles you can apply to now:

See all open AI jobs

This role is no longer accepting applications

Don't miss the next role like this

Create a free profile with your skills. We match you to new AI roles as they open, and companies hiring can contact you directly.

Create free profile, 2 minutes

Similar open roles you can apply to now:

See all open AI jobs

About the role

Customer problem: too much manual inspection and too much disagreement about borderline defects. Technical problem: model improvements that look good in one factory do not always survive another camera setup, another line speed, or another defect mix. Hiring problem: we need one senior person who can connect research-quality experimentation with production reality. If you have done applied vision in manufacturing, you probably already understand why this role exists.

Responsibilities

Lead experimentation on inspection models for high-speed glass and packaging lines
Design robust evaluation methods across plants, camera setups, and defect mixes
Work with applied engineers to bring promising model changes into production-ready form
Document experimental rationale and failure modes for internal and customer-facing review
Help shape the team's medium-term technical direction for inspection ML

Requirements

6+ years of computer vision or applied ML engineering
PyTorch for advanced model experimentation and training
Computer Vision experience in industrial inspection, manufacturing, or defect detection is strongly preferred
Docker for reproducible experimentation and deployment handoff
NumPy for image-processing workflows and error analysis
Experience evaluating models under distribution shift rather than only static benchmark conditions