About the Role

We are seeking an experienced Computer Vision Developer to design, build, and deploy production grade CV models that run reliably at the edge. You will own the full lifecycle — from model development through optimization, deployment on embedded and IoT hardware, and ongoing MLOps

— delivering real-time vision systems that operate in resource-constrained, real-world environments.

Job Location: Chennai/Bangalore (Hybrid)

Key Responsibilities

• Develop, train, and optimize production-grade computer vision models — object detection,

instance segmentation, classification, and tracking.

• Deploy and optimize models on embedded AI platforms and edge devices for low-latency, real

time inference.

• Build and integrate end-to-end CV pipelines into IoT systems and edge-to-cloud architectures.

• Apply model optimization techniques — quantization, pruning, distillation, and hardware

acceleration — to meet edge constraints.

• Establish and maintain robust MLOps practices: CI/CD for ML, model versioning, monitoring, drift

detection, and automated retraining pipelines.

• Benchmark and profile model performance (latency, throughput, memory, power) across target

hardware.

• Collaborate with hardware, firmware, and product teams to ship reliable, scalable edge AI

solutions.

• Maintain high standards of code quality, documentation, and reproducibility across the model

lifecycle.

Required Skills & Experience

• 6 – 8 years of hands-on experience in computer vision and deep learning.

• Strong proficiency in Python and CV / DL frameworks (PyTorch, TensorFlow, OpenCV).

• Proven track record building and deploying production-grade CV models.

• Deep expertise in instance segmentation and real-time CV systems.

• Hands-on embedded AI / edge deployment experience (e.g., NVIDIA Jetson, TensorRT,

OpenVINO, Coral, ARM-based platforms).

• Experience with IoT architectures and edge–cloud integration.

• Strong MLOps skills: CI/CD, containerization (Docker), model registries, monitoring, and

automated retraining.

• Solid understanding of model optimization for constrained hardware (quantization, pruning,

compression).

Nice to Have

• C++ for performance-critical inference paths.

• Familiarity with cloud ML workflows (AWS / Azure / GCP).

• Exposure to ONNX, GStreamer, or NVIDIA DeepStream pipelines.

• Knowledge of MQTT or other IoT messaging protocols.

• Experience with Kubernetes / Kubeflow for ML orchestration.

What We Offer

• Opportunity to work on cutting-edge edge AI and real-time vision products.

• A collaborative, hands-on engineering culture based in Chennai.

• Ownership of the full model lifecycle, from research to production.