Vision AI Engineer

Job Summary

We are looking for a Vision AI Engineer to deliver production-grade computer-vision solutions for industrial environments. In addition to model development, the role covers the image-acquisition system, including cameras, optics, lighting, positioning, triggering and calibration, so that models receive consistent and usable visual data.

Key Responsibilities

• Assess inspection, detection, safety or traceability requirements and translate them into measurable computer-vision acceptance criteria.
• Design image-acquisition setups covering camera type, lens, field of view, resolution, frame rate, lighting, mounting, triggering and calibration.
• Plan and execute data collection; define annotation guidelines and manage image/video dataset quality, coverage, versioning and governance.
• Develop and evaluate classification, object-detection, segmentation, tracking, OCR or anomaly-detection models as required.
• Optimize models for accuracy, latency, throughput and resource consumption on edge or cloud targets.
• Build preprocessing, inference and post-processing pipelines and integrate results with Digital Twin, MES, quality and workflow systems.
• Define testing across products, shifts, lighting conditions, camera drift, occlusions and realistic failure modes.
• Implement model and system monitoring, dataset feedback loops, retraining triggers and controlled release practices.
• Support site installation, commissioning, root-cause analysis and performance stabilization.
• Document hardware settings, calibration, model versions, test evidence and operating procedures.

Required Skills and Qualifications

• 5-8 years of experience in computer vision, machine learning or image-processing solutions, preferably in production environments.
• Strong Python and practical experience with OpenCV and deep-learning frameworks such as PyTorch or TensorFlow.
• Hands-on experience with detection, segmentation, classification, tracking, OCR and image-quality assessment.
• Understanding of industrial cameras, lenses, lighting, calibration, triggering, networking and edge-compute constraints.
• Experience curating and annotating datasets, selecting metrics and evaluating false-accept/false-reject trade-offs.
• Experience deploying optimized inference pipelines using containers, APIs and edge accelerators.
• Ability to troubleshoot the complete vision chain from image formation to model output and system action.
• Bachelors or masters degree in computer science, electronics, instrumentation, robotics or a related discipline.

Preferred Qualifications

• Experience with visual quality inspection, defect detection, PPE/safety monitoring, identification, counting or dimensional checks.
• Exposure to NVIDIA Jetson, TensorRT, ONNX, industrial vision SDKs or equivalent edge technologies.
• Knowledge of MLOps, active learning, synthetic data and Digital Twin integration.
• Experience operating vision systems under variable plant conditions and production-line cycle-time constraints.

Expected Delivery Outcomes

• Repeatable image acquisition and a governed, representative dataset.
• Validated vision model meeting agreed accuracy, latency and throughput criteria.
• Deployable edge/cloud pipeline with monitoring, calibration and support documentation.

Working Style and Stakeholder Engagement

• Work as part of a cross-functional delivery POD with manufacturing, engineering, data, cloud, enterprise application and change teams.
• Communicate complex technical topics in clear business language and maintain disciplined documentation, risk tracking and knowledge transfer.
• Support client workshops, design reviews, testing, deployment and adoption activities; travel to manufacturing locations when required.