Computer Vision Engineer

Proxgy is an Indian deeptech startup based in Gurugram that specializes in Industrial Internet of Things (IIoT). Founded in 2020 by Pulkit Ahuja, the company focuses on creating smart wearables and safety products designed to enhance the productivity and safety of the blue\-collar workforce

Key Products

The company operates on a three\-layered technology stack: industrial IoT hardware, software middleware, and a SaaS subscription layer.

● SmartHat: A sensor\-packed safety helmet featuring 360\-degree cameras, fall detection, gas leakage sensors, and real\-time GPS tracking.

● AirHat: A specialized industrial helmet with a multi\-layered filtration system, including HEPA and activated carbon filters, to provide clean air in hazardous environments.

● Lockator: A smart 4G lock for logistics that uses cellular, satellite, and mesh networks to track cargo and prevent tampering.

● AudioCube: A multilingual 4G\-enabled soundbox for UPI and digital payment notifications, supporting up to 24 languages.

● BirdBox: A portable, autonomous surveillance unit capable of operating in any weather condition for up to 8 days.

Role Overview

We are looking for a Computer Vision Engineer to develop and deploy AI\-based inspection systems for manufacturing use cases such as hinge detection, hole detection, missing component detection, and visual quality inspection.

The engineer will work on model development, camera stream processing, deployment, and support of production vision systems across edge devices and cloud/on\-premise environments.

• *Required Skills**

Computer Vision \& AI

• Python
• OpenCV
• PyTorch
• YOLO (v8/v11 preferred)
• Object Detection
• Image Processing
• Model Training and Evaluation
• Dataset Preparation and Annotation

Video \& Streaming

• RTSP Camera Streams
• GStreamer or FFmpeg
• Real\-time Video Processing
• Multi\-camera Stream Handling

Deployment

• Docker
• Linux
• GPU\-based Inference Deployment
• ONNX Runtime or TensorRT

Backend Development

• FastAPI or Flask
• REST APIs
• Basic Database Knowledge (PostgreSQL preferred)

Cloud / Infrastructure

• Basic AWS, Azure, or GCP experience
• Understanding of on\-premise and cloud deployments

Preferred Skills

• Industrial inspection systems
• Manufacturing or automotive domain experience
• NVIDIA Jetson deployment
• MQTT
• MLOps tools such as MLflow

Key Responsibilities

• Develop computer vision models for industrial inspection use cases.
• Build and optimize real\-time inference pipelines.
• Handle camera stream integration and processing.
• Deploy models on edge devices and servers.
• Support cloud and on\-premise deployments.
• Monitor model performance and improve accuracy.
• Collaborate with product and engineering teams for production rollout.

Ideal Candidate

• 2–5 years of hands\-on Computer Vision experience.
• Has deployed at least one vision model to production.
• Comfortable working with real\-time camera streams.
• Can independently handle model training, inference optimization, and deployment.
• Strong problem\-solving skills and ownership mindset.

Pay: ₹393,658\.13 \- ₹1,214,772\.27 per year

Benefits

• Flexible schedule
• Paid time off
• Provident Fund

Work Location: In person