Computer Vision Engineer - Real-Time Edge AI for Security

Laelaps; Zürich; Vor Ort; CHF 120.000 - 180.000; Vollzeit

Benefits dieser Stelle

Ownership in team build

Mission-driven work in security

Zusammenfassung

Laelaps AI, based in Zurich, is building intelligent perception for autonomous security. We seek a Computer Vision Engineer to own perception systems across multi-sensor stacks, starting from off-the-shelf models and training in-house when needed to deliver reliable real-time performance on edge hardware.

You will optimize detection, tracking, and re-identification across cameras, integrating vision-language models and ensuring low-latency streaming from cameras to models.

Qualifikationen

• Shipped computer vision or ML systems in production.
• Strong understanding of cross-camera reID, detection, and multi-object tracking.
• Hands-on with VLMs and real-time video streaming on edge hardware.

Aufgaben

• Design and ship cross-camera re-identification for people and vehicles across networks.
• Build high-throughput object detection and multi-object tracking pipelines for multiple streams.
• Integrate vision-language models for open-vocabulary detection and scene understanding.
• Own video ingestion and streaming path with attention to latency and resilience.
• Optimize and deploy models on edge hardware using TensorRT, quantization, and pruning.
• Evaluate and fine-tune existing models and train custom models when needed.
• Work with software and robotics teams to connect perception outputs to autonomy and alerting.

Kenntnisse

Computer Vision

Python

PyTorch

Edge computing

RTSP/WebRTC

Model deployment

TensorRT

ReID

MLOps

Tools

TensorRT

NVIDIA DeepStream

OpenCV

Jobbeschreibung

Laelaps AI, based in Zurich, is building intelligent perception for autonomous security. We seek a Computer Vision Engineer to own perception systems across multi-sensor stacks, starting from off-the-shelf models and training in-house when needed to deliver reliable real-time performance on edge hardware.

You will optimize detection, tracking, and re-identification across cameras, integrating vision-language models and ensuring low-latency streaming from cameras to models.