Find remote Computer Vision jobs.
Apply in minutes.

Create a free profile and browse the CV engineer positions on this page. Lead your profile with your specific domain (medical, automotive, retail, industrial) and deployment target (cloud, edge, embedded). Companies searching for CV engineers filter by domain — candidates who make these explicit in their headline and project descriptions consistently receive more responses than those who list frameworks without context.

Remote CV engineers earn $100K–$185K/year. Engineers specializing in autonomous systems or medical imaging are at the upper end. Edge AI specialists with TensorRT and embedded deployment experience are scarce and command premiums. Mid-level CV engineers (3–6 years, production deployment experience) typically earn $110K–$150K. Senior engineers who own full CV system design earn $150K–$185K+.

PyTorch Vision, OpenCV, YOLO variants, Detectron2 or MMDetection for detection, and the specific visual tasks you have delivered (classification, detection, segmentation, OCR). Include your evaluation metrics and the performance thresholds your systems needed to meet. For edge roles, add TensorRT, ONNX, Core ML, or OpenVINO experience. Always include your domain — medical, automotive, retail — as it is the primary filter companies use when searching for CV candidates.

Yes — entirely free. Creating a profile, browsing CV roles, and applying cost nothing. No placement fee, no revenue percentage, no exclusive contracts.

Yes, for most software-focused CV roles. Model training, pipeline development, evaluation, and architecture work are all effectively done remotely. The exception is roles that require hands-on work with physical hardware — camera calibration, embedded system setup, or on-site data collection. Most RemoteAI CV roles are software-focused and fully remote-compatible. If a role has any on-site requirements, the job description will specify it.

Start with model optimization fundamentals: quantization (INT8, FP16), pruning, and knowledge distillation. Then learn TensorRT for NVIDIA hardware or ONNX Runtime for cross-platform deployment. Get hands-on with a real edge device — NVIDIA Jetson Nano is a widely used starting point. Document a project where you took a full-precision model and optimized it for a constrained hardware target, measuring the accuracy vs. latency trade-off. This combination consistently gets interviews for edge CV roles.

Companies value CV engineers who understand the full data pipeline — not just model training on pre-cleaned benchmarks. If you have designed annotation guidelines, worked with annotation teams, built labeling pipelines (LabelStudio, CVAT, Scale AI), or curated training datasets from raw footage, describe this experience explicitly. It signals that you understand where models actually fail in production and that you can diagnose data quality issues independently.

The transition requires demonstrating production experience, not just research capability. Start by shipping a real project end-to-end: data collection, annotation, training, optimization, and deployment. Focus on an industry domain (retail, manufacturing, agriculture) rather than benchmark datasets. Document the engineering decisions — not just the model architecture — including the deployment environment, the performance constraints, and the data quality challenges. Companies hiring for industry CV roles want engineers who can navigate real-world constraints, not those who optimize for academic benchmarks.