Senior Computer Vision Engineer

About Aptly Automation
Aptly Automation is an innovative technology organization focused on solving complex real-world problems using advanced computer vision and data science. Our team members collaborate across disciplines to build scalable, production-ready systems that power intelligent products and services. We value curiosity, rigorous engineering practices, and continuous improvement in both research and development. Our remote team structure supports flexibility, inclusivity, and access to diverse talent. Joining us offers the opportunity to work with cutting-edge tools, contribute to impactful solutions, and grow in a supportive environment.
About the Role
We are looking for a Senior Computer Vision Engineer to design, implement, and optimize computer vision algorithms and systems. This is a full-time remote role. You will develop and deploy models for image and video analysis, image classification, object detection, tracking, and recognition, and run rigorous experimentation and evaluation. You will own solutions end to end, from early experimentation through production deployment, and work with cross-functional teams to define requirements, integrate vision components into larger software systems, and ensure reliable performance in production.
Key Responsibilities
Design, build, and train computer vision models for image and video analysis, image classification, object detection, tracking, and recognition.
Develop and optimize image processing, pattern recognition, and video-AI pipelines.
Build reproducible training pipelines with configuration management, hyperparameter optimization, and experiment tracking.
Evaluate model performance through sound experimental design, and improve accuracy through iterative data refinement, retraining, and evaluation cycles.
Export, validate, and optimize models for production (e.g., ONNX), balancing accuracy, latency, throughput, and infrastructure cost.
Build and maintain production CV services and APIs, and deploy them on cloud and edge platforms.
Carry solutions from prototype to production, owning them end to end across multiple business workflows.
Research and prototype new approaches to solve complex real-world problems.
Curate, annotate, and manage image and video datasets.
Collaborate with cross-functional teams to define requirements and translate them into technical solutions.
Review and improve existing code bases, and contribute to architecture decisions and best practices.
Review code, mentor team members, and promote engineering best practices.
Maintain clear technical documentation.
Contribute to the long-term technology strategy of the company.
Required Qualifications
Education: Bachelor’s degree (BS) in Artificial Intelligence, Computer Science, or a related field (Electrical Engineering, Data Science, etc.). A Master’s degree is a plus.
Experience: 5+ years of professional experience in computer vision and machine learning, including designing, training, and deploying models in production.
Programming: Strong command of Python, with solid software engineering fundamentals (algorithms, data structures, clean and testable code). C++ is a plus.
Deep learning frameworks: Hands-on experience with PyTorch (required) and related tools such as torchvision and PyTorch Lightning. TensorFlow/Keras is a plus.
Image classification, detection, and segmentation: Practical experience with image classification (CNNs, transfer learning) and object detection, including YOLOv8 and the Ultralytics framework (training, fine-tuning, optimization, and deployment), and similar architectures such as Faster R-CNN, Mask R-CNN, RT-DETR, and U-Net. Familiarity with Detectron2, MMDetection, or Segment Anything (SAM) is a plus.
Video AI and tracking: Experience building video-AI pipelines and multi-object tracking methods (e.g., ByteTrack, DeepSORT).
Computer vision libraries: Proficiency in OpenCV, scikit-image, and Pillow, and classical image processing techniques such as feature extraction (SIFT, ORB), filtering, morphology, and edge and contour detection.
ML and DL foundations: Strong understanding of machine learning and deep learning, including CNNs, Vision Transformers, transfer learning, loss functions, optimization, and regularization.
Data and evaluation: Experience handling and augmenting image and video datasets (Albumentations, torchvision transforms, Kornia), plus rigorous model evaluation (precision, recall, mAP, IoU, confusion analysis) and experimental design.
MLOps and reproducibility: Experience building reproducible training pipelines and tracking experiments (e.g., Hydra, Optuna, MLflow, Weights & Biases).
Model serving: Experience exporting models to ONNX and serving them through production APIs (FastAPI or similar) with inference optimization.
Data science toolkit: NumPy, Pandas, scikit-learn, Matplotlib/Seaborn, and Jupyter for analysis and experimentation.
Language: Excellent English proficiency, written and spoken, for clear communication with cross-functional teams.
Remote work: Comfortable working remotely and independently, with strong self-management and communication skills.
Problem solving: Strong analytical and problem-solving skills and the ability to own projects end to end.
Preferred Qualifications
Deployment and optimization: ONNX Runtime, TensorRT, OpenVINO, TensorFlow Lite, NVIDIA Triton Inference Server, model quantization and pruning, performance profiling, CUDA fundamentals, and edge deployment.
Serving and APIs: FastAPI (async), Flask, Pydantic, and REST API design for inference services.
MLOps: Docker, Kubernetes, Hydra, Optuna, MLflow, Weights & Biases, TensorBoard, DVC, and CI/CD with GitHub Actions.
Cloud and GPU: AWS (SageMaker), GCP (Vertex AI), or Azure ML, plus GPU-based and distributed training.
Annotation tools: CVAT, Roboflow, Label Studio, or LabelImg.
Advanced vision tasks: Instance segmentation, anomaly detection, OCR and document understanding, and exposure to Hugging Face Transformers, timm, vision-language models, and generative models such as diffusion models.
Collaboration: Experience with version control, collaborative development workflows, and multidisciplinary teams.
Leadership: Prior experience mentoring engineers or leading technical projects.
Tech Stack
Languages: Python, C++, SQL
Deep learning: PyTorch, PyTorch Lightning, TensorFlow/Keras, torchvision, timm, Hugging Face Transformers
Detection, tracking, and segmentation: YOLOv8, Ultralytics, Detectron2, MMDetection, SAM, ByteTrack, DeepSORT
Computer vision: OpenCV, scikit-image, Pillow, Kornia, Albumentations
Data science: NumPy, Pandas, scikit-learn, Matplotlib, Seaborn, Jupyter
Optimization and deployment: ONNX, ONNX Runtime, TensorRT, OpenVINO, TensorFlow Lite, Triton, CUDA
Serving and APIs: FastAPI, Flask, Pydantic, REST APIs
MLOps and tooling: Docker, Kubernetes, Git, Hydra, Optuna, MLflow, Weights & Biases, TensorBoard, DVC, GitHub Actions
Annotation: CVAT, Roboflow, Label Studio
Cloud: AWS (SageMaker), GCP, Azure, cloud GPU environments
What We Offer
Fully remote, full-time position
Opportunities to work with cutting-edge tools and contribute to impactful solutions
A supportive environment focused on curiosity, rigorous engineering, and continuous improvement
Opportunities to lead technical strategy and mentor a growing team

Work arrangement
Yes