Midhun P M — Case studyETLab+

The problem

Same enemy as EtlabPro: the ETLAB portal every Sahrdaya student fights daily. ETLab+ was my first shot at it, back in semester three — before I knew what a service layer was or why you shouldn't let clients trigger scrapes.

What I built

A cross-platform mobile app in React Native (Expo) with a Spring Boot backend, distributed straight to classmates as an APK:

• JWT-based auth against the backend.
• Attendance, timetable, and results — including end-semester results — pulled from ETLAB with real-time scraping.
• An analytics dashboard for the numbers students actually care about.
• AI query answering — ask questions about your own academic data in plain language, backed by a separate AI API service I built alongside it.

It won Best S3 Project.

Tech decisions

Spring Boot for the backend. I wanted to learn Java properly, and nothing teaches you structure like Spring. Controllers per domain — auth, attendance, timetable, results — JWT secret and expiry in environment config, CORS locked down, and a Dockerfile so deployment was one command.

Expo for the app. React Native let me reuse what I knew from React, and Expo made APK builds and distribution to classmates painless. No Play Store review queue, just a file people install.

What I learned

This project taught me that "it works on my phone" and "it works for fifty students" are different engineering problems. Real users on real networks found every unhandled edge: expired sessions, portal timeouts, stale data presented as fresh.

Everything ETLab+ taught me the hard way — sync policies, normalization, separating the scraper from the API surface — became the foundation for EtlabPro, its successor. This is the project I grew up on.

Links

• GitHub