Hi, I'm Karuppasamy

A production-grade web application used to inspect solar plants and transmission lines using aerial imagery and geospatial data.

• Processes images, videos, IR & RGB tiles (5GB–25GB)
• Interactive GIS map visualization using Leaflet
• Asynchronous processing with Celery & RabbitMQ
• Role-based access: Admin, Internal, Client

Tech Stack:

Django, REST API, PostgreSQL, Leaflet, Pandas, AWS (EC2, S3, RDS, Route 53,codecommit)

Team Size: 3

Deployment: AWS Production (Handled by me)

An internal enterprise system for managing users (IAM), expenses, and inventory operations.

• User & role management with JWT authentication
• Expense tracking and inventory modules
• React frontend using Context API & Axios

Tech Stack:

Django, React.js, PostgreSQL, REST API, AWS (s3,codecommit)

Team Size: 8

Deployment: Internal (Handled by another team member)

A real-time IoT-based motion detection system built using RCWL-0516 microwave sensor and Raspberry Pi Zero 2 W. The system captures images on motion detection and reliably syncs data with the server, even during network failures.

• Case 1 – Normal Connectivity:
  On motion detection, images are captured and stored locally, then sent to the server via REST API. Once the server responds successfully, the local images are automatically deleted from the Raspberry Pi to save storage.
• Case 2 – Network Failure Handling:
  If server connectivity is unavailable during motion detection, images are safely stored locally. When the connection is restored, all pending images are automatically sent to the server and removed locally after successful upload.
• Case 3 – Device Health Monitoring:
  Implemented device health checks and Last Will & Testament (LWT) using RabbitMQ (MQTT protocol) to detect Raspberry Pi online/offline status in real time.

Tech Stack:

Python, Flask, REST API, MongoDB, RabbitMQ (MQTT), Raspberry Pi Zero 2 W, RCWL-0516, HTML, CSS, JavaScript,GitLab

Repository:Iot Cloud(Private)