Are you passionate about turning complex visual data into real-world AI impact? At Digital Place Vision, we are on a mission to transform the retail industry using cutting-edge computer vision to tackle planogram compliance and retail shelf analytics at scale (recognizing over 5,000+ SKUs!).
We’re looking for a Senior Computer Vision Engineer to design, build, and deploy end-to-end CV systems—from edge devices (Android) to cloud inference.
💡 What You’ll Do:
– End-to-End CV Systems: Lead the design of AI pipelines for product detection, recognition, localization, and shelf segmentation.
– Model Optimization & Edge AI: Optimize models for inference acceleration on Android/Edge devices and cloud environments (TensorRT, ONNX Runtime, CUDA).
– Real-World Field Testing: Conduct extensive testing in actual retail/warehouse environments to solve tricky edge cases (lighting, camera angles, motion blur).
– MLOps & Observability: Establish production-ready MLOps workflows, tracking model drift, metrics, and deployment pipelines.
🛠️ What We’re Looking For:
– Tech Stack: Strong proficiency in Python, C++, PyTorch / TensorFlow, and OpenCV.
– Edge & Cloud Deployment: Hands-on experience optimizing models for compute-constrained environments (Android GPU/NCNN/TensorRT) and cloud platforms (GCP/AWS/Azure, Docker, K8s).
– Tooling: Experience with dataset analysis & MLOps tools (FiftyOne, MLflow, W&B, TensorBoard, Grafana).
– Bonus Points: Experience with ROS / ROS2, barcode decoding libraries (Halcon, ZBar), or retail/logistics CV systems!
🎁 What We Offer
– Build real, cutting-edge AI robotics products deployed directly into commercial settings.
– Lead core visual algorithm architecture from the ground up.
– Collaborate deeply with top-tier AMR and drone R&D teams.
– Competitive compensation & an innovation-first culture.
🌐 Learn more about us: digitalplace.ai
🔗 Follow us on LinkedIn: Digital Place Vision Pte Ltd
📥 Interested? Send your updated CV to Form
Let’s shape the future of smart retail together. 🤖✨
This contact form is available only for logged in users.