Rumeasiyan:
Suseenthiran Arulraj RumeasiyanFull-Stack Developer & ML Enthusiast
Rumeasiyan
Suseenthiran Arulraj Rumeasiyan
Undergraduate, Computer Science
Colombo, Sri Lanka

Table of Contents
Abstract
I'm a computer science undergraduate based in Sri Lanka, building full-stack web applications with Laravel and Next.js, and experimenting with applied machine learning in Python. I like projects where real-world data meets clean product UX: dashboards, automation tools, and AI-assisted workflows. Recently, I built an AI weather companion that recommends location-aware activities using real-time weather APIs and Gemini, and a spoiler-blocking browser extension that classifies social content and blurs spoilers with user feedback loops. I'm actively seeking projects where I can contribute as a developer while sharpening system design and software architecture skills.
Keywords:Laravel · Next.js · Tailwind CSS · MySQL · Python · Gemini API · DevOps · Security · Machine Learning
1. Introduction & Highlights
The primary focus of this site is to demonstrate the practical application of full-stack engineering principles combined with modern AI capabilities. Key contributions include:
- Integration of LLMs in Consumer Apps: Utilizing the Gemini API to provide context-aware data processing in real-time environments.
- Privacy-Centric Design: Implementing local-first filtering where possible before offloading to cloud inference.
- Scalable Architecture: Leveraging Laravel's ecosystem and Next.js for high-performance frontend delivery.
2. Selected Works
2.1 WeatherCraft AI Companion (2024–2025)
Problem: Tourists need alternative activity plans when weather changes unexpectedly, but static weather apps lack context-aware recommendations.
Methods: Implemented a real-time weather tracking system coupled with Gemini Pro to generate dynamic activity suggestions based on current meteorological conditions and user location.
2.2 NPZ Spoiler Blocker (2024)
Problem: Social media feeds often contain spoilers for trending media, ruining the experience for users who haven't caught up.
Methods: Developed a browser extension utilizing a keyword filter followed by LLM-based semantic classification to identify and blur potential spoilers. Includes a user feedback loop inspired by RLHF principles to improve accuracy.
Results: Achieved high precision in spoiler detection with low latency impact on page load.
3. Experience
Computer Science Undergraduate
Focusing on software engineering, artificial intelligence, and data structures. Active member of the university tech society.
4. Skills Graph
Frontend
- React
- Next.js
- Tailwind CSS
- TypeScript
Backend
- Laravel (PHP)
- Node.js
- MySQL
- Python
AI & ML
- Gemini API
- Prompt Engineering
- Scikit-learn
- Pandas
Tools
- Git
- Docker
- Figma
- Linux
5. References & Contact
This site serves as a living document of my technical journey. I am currently open to collaboration on projects involving full-stack development and applied machine learning.
- Email: rumeasiyan@gmail.com
- GitHub: github.com/Rumeasiyan
- LinkedIn: linkedin.com/in/rumeasiyan
@online{Rumeasiyan2026,
author = {Rumeasiyan, Suseenthiran Arulraj},
title = {Rumeasiyan: Personal Website},
year = {2026},
url = {https://github.com/Rumeasiyan},
note = {Accessed: 2026-03-29}
}Cite this site if you reuse the code structure or visual system.
Appendix A: Other Projects
- A.1Analytics Dashboard for E-commerce
- A.2Personal Finance Tracker (Income/Expense)
- A.3Restaurant Booking System
- A.4Sales Forecasting Model (Python)
- A.5Sentiment Analysis Tool on Twitter Data