U
UnitX Tech Stack
Deep learning vision system for automated defect inspection on manufacturing lines
UnitX deploys computer-vision robots for visual inspection tasks on production lines, built on PyTorch, TensorFlow, and ROS 2 with physics simulation via Isaac Sim and Gazebo. The tech stack and active projects reveal a company focused on closing the sim-to-real gap—building synthetic data pipelines and real-time inference engines (<20ms latency) to move models from simulation into factory floors. Hiring is accelerating across engineering and HR, signaling both technical scaling and early organizational pain as they grow beyond startup structure.
Tech Stack 22 technologies
What UnitX Is Building
◆Challenges
- Repetitive factory-floor problems
- Scaling hr function
- Building middle manager layer
- Engaging global distributed team
- Low-latency defect detection
- Automating repetitive visual tasks
- Ensuring quality in manufacturing
- Scaling ai to global factories
- Closing sim-to-real gap
- Reducing synthetic data pipeline bottlenecks
▲Active Projects
- End-of-line large field-of-view line
- Robotics manipulation products in manufacturing
- Build middle manager layer
- Run engagement surveys
- Own hr ops backbone
- Defect segmentation on high-resolution images
- Real-time inference pipeline under 20ms
- Model performance metrics and tools
- Physics simulation environments
- Sim-to-real data flywheel
Hiring Activity
9 roles · 7 in 30d
Department
Seniority
Company intelligence
