Senior Computer Vision Engineer

About Us

We are an internationally oriented company dedicated to creating intelligent technologies, digital solutions, and data-driven products that enhance operational efficiency, customer satisfaction, and overall business performance across various industries. Our multidisciplinary teams work collaboratively across AI, machine learning, software engineering, data science, product development, robotics, automation, operations, and research to design scalable computer vision systems that transform visual data into actionable insights.

Job Overview

As a Senior Computer Vision Engineer, you will spearhead the design, development, optimization, deployment, and ongoing enhancement of cutting-edge computer vision and image-processing applications. Your role involves leveraging expertise in deep learning, machine learning, image processing, and software engineering to build robust production-grade systems addressing tasks like object detection, image classification, segmentation, tracking, recognition, video analytics, and visual inspection.

Key Responsibilities

• Architect and implement sophisticated computer vision systems tailored to practical use cases.
• Lead all project phases from defining problems and preparing data to developing, validating, deploying, and optimizing models.
• Translate operational and business needs into concrete computer vision solutions and technical designs.
• Examine images, videos, sensors, and multimodal data to identify suitable computer vision methods.
• Create pipelines for image processing and computer vision tasks.
• Develop models for object detection, various segmentation techniques, tracking, pose estimation, and recognition.
• Build video analytics solutions for both real-time and offline environments.
• Employ deep learning architectures including CNNs, transformers, and vision transformers appropriately.
• Choose and implement appropriate models, algorithms, architectures, and frameworks.
• Use PyTorch, TensorFlow, OpenCV, or similar frameworks for modeling and optimization.
• Perform image preprocessing steps including augmentation, normalization, enhancement, filtering, feature extraction, and transformation.
• Create methods to manage real-world challenges like lighting variability, weather conditions, occlusion, motion blur, and noise.
• Design and manage data pipelines for collection, cleaning, labeling, validation, and dataset maintenance.
• Set standards for image/video annotation and ensure quality control and dataset governance.
• Collaborate with labeling teams and external annotators to enhance training data quality.
• Analyze datasets for coverage, class imbalances, biases, edge cases, and distribution variances.
• Develop strategies including dataset expansion, augmentation, synthetic data generation, and mining hard examples.
• Conduct model training, fine-tuning, and evaluation.
• Systematically optimize hyperparameters and experiment with models.
• Define evaluation metrics suited to specific computer vision tasks.
• Analyze model performance metrics such as precision, recall, F1 score, IoU, mAP, ROC-AUC, latency, and throughput.
• Investigate error patterns including false positives/negatives and misclassifications.
• Benchmark models against alternative architectures and datasets.
• Optimize computational performance for speed, memory, efficiency, and scalability.
• Deploy models on cloud platforms, edge devices, embedded systems, or specialized hardware.
• Build real-time inference pipelines for imaging and video applications.
• Utilize optimization techniques like quantization, pruning, distillation, batching, and hardware acceleration where beneficial.
• Work with diverse computing platforms including GPUs, CPUs, edge accelerators, and embedded systems.
• Integrate models with APIs, applications, robotics, automation, databases, and enterprise software.
• Partner with software engineering teams to deliver reliable, production-grade computer vision services.
• Develop scalable APIs, inference services, and data pipelines along with supporting software.
• Implement model version control, deployment, monitoring, rollback, and lifecycle management.
• Set up monitoring to detect model drift, degradation, performance shifts, and anomalies.
• Automate testing and validation routines for computer vision systems.
• Establish reproducible workflows for machine learning and computer vision development.
• Document models, datasets, architectures, experiments, assumptions, limitations, and deployment details comprehensively.
• Work with product and operations teams to define measurable success criteria and outcomes for solutions.
• Conduct proofs-of-concept, feasibility studies, and prototype validations.
• Support pilot projects and transition validated prototypes to production.
• Collaborate with robotics and automation teams on perception, recognition, localization, and scene understanding systems.
• Support applications across autonomous systems, industrial inspection, warehouse automation, security, retail analytics, mobility, and intelligent infrastructure.
• Stay abreast of emerging technologies including foundation models, multimodal models, generative AI, and vision-language models.
• Evaluate practical uses for AI innovations such as synthetic data and multimodal AI approaches.
• Monitor research trends, open-source projects, industry benchmarks, and novel computer vision methodologies.
• Assess third-party platforms, models, APIs, hardware, and vendor technologies.
• Manage collaborations with external technology vendors, researchers, consultants, and experts when needed.
• Contribute to intellectual property creation, technical research, publications, patents, and innovation initiatives.
• Uphold responsible AI principles emphasizing privacy, fairness, transparency, security, and ethical use of visual data.
• Ensure solutions comply with data protection, cybersecurity, regulatory, and organizational standards.
• Mentor and guide junior engineers, machine learning engineers, and data scientists.
• Review technical designs and provide architecture and model development guidance.
• Set standards for computer vision engineering, development practices, and reusable frameworks.
• Inform leadership regularly about project status, model performance, risks, progress, and innovation prospects.

Ideal Candidate Profile

• Extensive experience in computer vision, deep learning, machine learning, image processing, AI, robotics perception, video analytics, or visual intelligence domains.
• Background in technology sectors such as robotics, manufacturing, automotive, logistics, retail, security, or healthcare is preferred.
• Strong grasp of computer vision fundamentals and contemporary deep-learning techniques.
• Proven track record in deploying production-grade computer vision systems.
• Expertise in image processing, feature extraction, detection, segmentation, classification, and tracking.
• Proficiency with deep learning models including CNNs, transformers, and vision transformers.
• High competence in Python programming and industry-standard ML frameworks.
• Hands-on experience with PyTorch, TensorFlow, OpenCV, or equivalent tech stacks.
• Strong capabilities in model training, evaluation, optimization, and deployment.
• Experience managing large-scale image and video datasets.
• Knowledge of data annotation, dataset curation, augmentation, and quality assurance.
• Ability to analyze model errors and enhance edge-case performance.
• Familiarity with model optimization methods such as quantization, pruning, distillation, or hardware acceleration is highly advantageous.

Performance Indicators

• Model accuracy and object detection mean average precision (mAP).
• Metrics including precision, recall, F1 score, Intersection over Union (IoU), and classification accuracy.
• Segmentation quality and tracking accuracy.
• False-positive and false-negative rates.
• Inference latency and real-time frame processing rates.
• GPU/compute utilization and memory efficiency.
• Deployment success and production availability of models.
• Rates of model degradation and data-drift detection.
• Model monitoring coverage and system reliability.
• Quality and completeness of datasets and annotations.
• Training and experimentation cycle durations.
• Prototype to production implementation success rate.
• Project delivery timeliness and incident rates.
• Frequency of model rollback and reduction of technical debt.
• Automated testing proportion and documentation completeness.
• Stakeholder satisfaction and innovation contributions.
• Security and privacy compliance adherence.
• Team growth and mentoring effectiveness.