Computer Vision Engineer - Python / OpenCV

Role : Computer Vision Engineer Edge Deployment (NBFC : Hoodi, : 7 to 9 : Financial Services / NBFC

About the Role :

We are looking for a Computer Vision Engineer to lead and scale vision-based systems for our smart asset-monitoring and physical verification products. This is a high-impact role where your work will directly contribute to building proprietary visual intelligence tools that support field verification, customer onboarding, and fraud prevention all from the edge.

This position combines deep expertise in vision algorithms, embedded optimization, and real-world system deployment. Youll be part of a cross-functional team focused on bringing scalable, intelligent solutions to operational challenges across our NBFC ecosystem.

Key Responsibilities :

• Lead vision data collection efforts in real-world settings including setting up annotated video datasets under variable lighting and environment conditions.
• Design and implement computer vision pipelines using Python, OpenCV, and deep learning frameworks (PyTorch / TensorFlow) for detecting visual cues such as color shifts, reflections, lighting variations, etc.
• Work extensively with 3D and RGB data to analyze patterns and identify objects or abnormalities with precision.
• Deploy models to ARM-based SoCs, optimizing for power consumption, latency, and CPU / GPU constraints on edge devices.
• Invent, iterate, and benchmark novel computer vision algorithms, especially for scenarios with no existing precedents in the BFSI domain.
• Apply traditional DSP techniques (Fourier, SVD, PCA, optical flow) and linear algebra for preprocessing, filtering, and enhancement.
• Collaborate with hardware and software teams to ensure reliable on-device inference, including model quantization and compression for low-power inference.

Must-Have Skills :

Good to Have :

(ref : hirist.tech)