Role Summary

We are looking for a hands-on Senior Computer Vision Engineer to lead the development of a rapid Proof-of-Concept (POC) vision system.

The hardware platform is not yet finalized, so the ideal candidate must be comfortable building hardware-agnostic, modular computer vision pipelines that can later be adapted for edge deployment.

This role is focused purely on CV model development, real-time inference, and pipeline optimization rather than frontend or backend development.

Key Responsibilities

• Design and implement end-to-end computer vision pipelines.

• Develop rapid POCs for object detection, tracking, segmentation, or OCR.

• Build real-time video processing systems.

• Optimize models for edge/embedded deployment (CPU/GPU/NPU).

• Convert and deploy models using ONNX, TensorRT, TFLite, or similar frameworks.

• Benchmark and improve performance metrics, including latency, FPS, and memory usage.

• Work with limited or synthetic datasets during early prototyping.

Required Skills

• Strong expertise in Python (must-have).

• Solid experience with C++ for performance optimization.

• Hands-on experience with PyTorch or TensorFlow.

• Strong knowledge of OpenCV.

• Experience with object detection frameworks (e.g., YOLO, SSD, Faster R-CNN).

• Experience building real-time inference pipelines.

• Knowledge of model optimization techniques such as quantization, pruning, and profiling.

Nice to Have

• Experience with edge devices (e.g., NVIDIA Jetson, ARM, Qualcomm).

• Experience in smart glasses, AR/VR, robotics, or embedded vision.

• Experience deploying models to ONNX Runtime or TensorRT.

What We’re Looking For

• Ability to work independently in an ambiguous environment.

• A strong problem-solving mindset.

• Fast prototyping and iteration capabilities.

• Proven experience delivering working POCs under tight timelines.