Perception Engineer (Computer Vision & Edge AI)

About the Company

This role is with a dynamic AI robotics startup focused on developing cutting-edge intelligent software to power next-generation autonomous physical security systems. Their mission is to enable robots and smart sensors to undertake critical and hazardous security tasks by integrating advanced AI, robotics, and computer vision technologies. Supported by experienced founders and top-tier investors, they create production-ready AI systems designed for real-world applications rather than research demonstrations. Joining this team offers a chance to be part of one of Europe's most exciting robotics ventures at an early stage and directly influence products deployed in operational environments.

Location: Zürich, Switzerland (On-site)

Employment Type: Full-time

Seniority Level: Mid-Senior

Visa Sponsorship and Relocation: Available across Europe

Role Overview

We are seeking a hands-on Perception Engineer responsible for developing production-quality computer vision systems that drive intelligent robotic platforms. The focus includes designing, deploying, and optimizing real-time perception pipelines covering object detection, multi-object tracking, cross-camera re-identification, video streaming infrastructures, edge AI performance enhancement, and modern vision-language model integration. Close collaboration with robotics, AI, and software engineering teams will ensure reliable real-world system performance.

Key Responsibilities

• Build and maintain production object detection, tracking, and cross-camera re-identification systems.
• Develop and optimize multi-camera perception pipelines for high performance.
• Incorporate modern Vision-Language Models to enhance scene understanding and open-vocabulary detection.
• Deploy and optimize deep learning models specifically for edge devices such as NVIDIA Jetson, using TensorRT, quantization, pruning, and latency optimizations to meet real-time requirements.
• Design and implement scalable video streaming pipelines using RTSP and WebRTC protocols, managing multi-stream processing, synchronization, latency, and reliability.
• Construct efficient end-to-end video ingestion and processing systems.
• Evaluate and fine-tune state-of-the-art open-source and commercial machine learning models; train custom models as needed.
• Establish robust data pipelines for image and video datasets, including reproducible evaluation and performance monitoring workflows.
• Collaborate with interdisciplinary teams to integrate perception functionalities into autonomous robotic systems and influence long-term perception architecture decisions.

Required Qualifications

• Minimum of three years experience in developing production computer vision or machine learning solutions.
• Proficient in Python programming and experienced with PyTorch or JAX frameworks.
• Strong background in object detection and multi-object tracking techniques.
• Experience deploying ML models on edge computing hardware.
• Hands-on expertise with TensorRT optimization and acceleration of machine learning models.
• Knowledge of quantization, latency reduction, and model pruning strategies.
• Familiarity with video streaming technologies such as RTSP and WebRTC.
• Proven ability to build scalable image and video data pipelines.
• Solid software engineering skills with a track record of taking systems from prototypes into production.

Preferred Skills

• Experience with cross-camera re-identification.
• Knowledge of Vision-Language Models (VLMs).
• Familiarity with NVIDIA DeepStream SDK.
• Expertise in MLOps, continuous integration/deployment for machine learning, and model monitoring.
• Background in robotics, autonomous systems, drone or surveillance technologies.
• Working knowledge of retrieval-augmented generation (RAG), large language model integration, and ROS2 framework.
• Academic contributions such as publications at conferences like CVPR, ICCV, NeurIPS, or ICLR are advantageous.

What We Provide

• Competitive salary package.
• Visa sponsorship and assistance with relocation across Europe.
• Opportunity to collaborate with leading AI, Robotics, and Physics experts.
• Access to extensive computational resources and authentic robotic datasets.
• High level of ownership and responsibility from the start.
• Small and agile engineering team fostering minimal bureaucracy.
• Chance to build AI systems actively used in real-world security settings.
• Modern office facilities located in Zürich.
• Inclusive and cooperative international engineering culture.

Hiring Process

Applications are reviewed by an AI-based matching system to identify suitable roles within our network, potentially unlocking multiple opportunities with one submission. All verified openings represent active recruitment efforts. Candidates with strong alignment will be supported through the hiring process, while others may receive alternative considerations without being neglected.