I'm a B.Tech AI & Data Science student at Ramco Institute of Technology. I build computer-vision systems, ML models and full-stack web apps, and I put them online.
Projects: PPE Detection, Exam Proctoring, Financial Management, Pulse ERP, Sahakar Sathi, Marine Oil Spill Detection, RIT ERP — AI Integration.
What I build
Real-time detection with YOLOv8, OpenCV and MediaPipe, applied to PPE compliance and exam integrity.
Data preprocessing, model training and evaluation in Python, then wrapped in interfaces people can use.
React frontends, Node.js/Express and Flask APIs, and MongoDB and SQL data layers.
Projects shipped to live URLs on Vercel and Render, with Docker and Dokploy in the toolkit.
Section: mathan/real-world
Ramco Institute of Technology
I don’t only build standalone student projects. I’ve also contributed AI integration work to the ERP platform at Ramco Institute of Technology.
The ERP is a larger institutional platform. My contribution is on the AI integration side, not the system as a whole.
Section: mathan/work
Computer vision, full-stack products, a hackathon build and real institutional platforms. Every entry opens a full case study.
07 entries
A computer-vision application that detects helmets, safety vests and masks in real time and generates PPE violation alerts.
TrustMeter monitors online exams with face and object detection, anomaly monitoring and secure reporting, built on Python, OpenCV, YOLO and Flask.
A responsive expense tracker with a dashboard, transaction management and spending insights, built with React, REST APIs, Node.js/Express and MongoDB.
Implemented portal features, resolved bugs, and improved UI components and backend functionality on the live Pulse portal.
An agriculture-focused web platform, developed and deployed for practical use, that helps people navigate cooperative services and schemes.
A satellite-based application to identify and monitor oil spills in marine environments, built and deployed for Smart India Hackathon 2026.
Contributed to AI integration within the ERP ecosystem at Ramco Institute of Technology: real institutional software, not a standalone demo.
Section: mathan/experience
AI Integration Contributor · RIT ERP
Contributed to AI integration within the ERP ecosystem at Ramco Institute of Technology, working inside a live institutional platform rather than a standalone project.
ML / DL Intern
Preprocessed data, and trained and evaluated ML/DL models in Python.
MERN Stack Intern
Built frontend and backend features with MongoDB, Express.js, React.js and Node.js.
Artificial Intelligence Intern
Developed and tested application features in an online internship.
DeploymentDeployed two real-world web applications; deployment tools include Docker and Dokploy.
Section: mathan/journey
From an AI & Data Science degree to detection systems, full-stack products, three internships and real institutional software.
Chapters, not a strict chronology. Dates appear where they’re on record.
Ramco Institute of Technology, Rajapalayam.
Machine learning, deep learning and computer vision, alongside certifications: Machine Learning with Python (IBM), Deep Learning Fundamentals (Coursera), AI Foundation (Microsoft Learn) and Deloitte's Data Analytics Job Simulation.
A PPE detection system on YOLOv8 and OpenCV, and TrustMeter, an AI-based exam proctoring system with face, gaze and object detection.
A financial management app on React, REST APIs, Node.js/Express and MongoDB, plus feature and fix work on the Pulse portal.
Developed and tested application features in an online internship.
Built frontend and backend features with MongoDB, Express.js, React.js and Node.js.
Preprocessed data, and trained and evaluated ML/DL models in Python.
Finalist at a National Level AI Hackathon, and 1st Prize at the Project Expo at Francis Xavier College.
A satellite-based marine oil spill detection and attribution console, built and deployed as a team.
Contributed to AI integration within the ERP ecosystem at Ramco Institute of Technology.
Sahakar Sathi and the oil-spill console live on Render, and the financial app on Vercel, with Docker and Dokploy in the deployment toolkit.
Section: mathan/stack
Technologies I’ve used across projects and internships, grouped like a periodic table. Pick a group to isolate it. (This site itself is just Next.js.)
Also on the resume
Principles
Before picking a model or a framework, I work out what the system actually needs to do and who it's for.
Get a working version running, then iterate. Complexity has to earn its place.
A detector only matters once it's wired into an API, an interface and a deployment. I work across both layers.
Projects go to live URLs, because deployment surfaces problems a local demo never will.
Internships in ML/DL, MERN and AI keep me moving between data, models and the web stack, and each one feeds the next build.
Section: mathan/lets-talk
I’m open to internships and to AI, computer-vision and full-stack work. The quickest way to reach me is below.
