<aside>
🧠
Operating rooms are visually complex, dynamic environments with strict reliability and privacy requirements. This role sits at the intersection of raw camera data, computer vision models, and clinical workflows — turning challenging visual input into robust, deployable perception that works in real surgical settings.
</aside>
Working at Lynqo
We build a software platform that is used in real hospital environments, with real constraints and real users. That means clarity, reliability and good communication matter just as much as technical elegance.
We value ownership, curiosity and openness. People at Lynqo take responsibility for their work from idea to deployment, invest time in understanding the context they’re building for, and aren’t afraid to ask questions or challenge assumptions.
We’re a small but growing team with short feedback loops. You’ll work closely with both technical and non-technical colleagues, and your input will shape how the product evolves.
<aside>
Your role in a nutshell
- Full-time role
- Close collaboration with our CTO, engineering team and hospital partners
- High ownership over core computer vision components, from model design to deployment
- Regular visits to partner hospitals for observation, testing and validation
- Flexible working setup (Antwerp office + remote)
</aside>
Your mission ahead
Perception & tracking
- Develop and optimise detection, segmentation and tracking models for complex operating room environments
- Improve accuracy under challenging conditions, including occlusions, partial visibility and dynamic scenes
- Build reliable vision pipelines that work consistently across procedures and hospital settings
Training & model quality
- Build reproducible training and evaluation pipelines
- Develop visual recognition methods for challenging objects
- Analyse failure modes and improve model accuracy across procedures and hospital environments
Edge deployment & robustness
- Optimise and deploy vision components on edge devices
- Balance accuracy, latency and robustness under real-world constraints
- Monitor and validate model performance in deployed systems