HostelSOS: AI-Powered Hostel Issue Management

Inspiration

Hostel life should feel safe, clean, and comfortable — but in reality, students often face leaking taps, broken lights, unhygienic bathrooms, poor food quality, and delayed responses from wardens. Existing hostel ERPs don’t solve the real issue: students are scared to complain because their identity is exposed.

That’s where the idea of HostelSOS was born — a simple, anonymous, AI-powered solution that gives students a voice without fear.

⭐ What it does

HostelSOS is an Arm-optimized, AI-powered mobile app that lets hostel students:

Report problems anonymously

Take photos → AI classifies the issue (electrical, plumbing, hygiene, food, etc.)

Automatically generate complaint tickets

Track the issue status in real-time

Receive updates as the hostel staff works on the problem

View proof of resolution uploaded by management

It protects privacy while ensuring accountability.

⭐ How we built it

We built HostelSOS as a mobile-first AI application optimized for Arm-based Android devices.

React Native for cross-platform mobile UI

Firebase (Auth, Firestore, Storage) for backend

TensorFlow Lite / ONNX Runtime Mobile for on-device AI

AdMob for monetization

Custom Node.js Cloud Functions for automated notifications and ticket management

The AI model is trained to classify common hostel problems using images and runs efficiently on Arm CPU/GPU hardware for fast, low-latency inference.

⭐ Challenges we ran into

We faced several challenges during development:

Getting the AI model to accurately classify hostel issues with limited training data

Optimizing the model to run on-device for Arm processors without slowing the app

Maintaining true anonymity while still allowing students to track ticket progress

Designing a simple UI suitable for stressed, rushed students

Ensuring Firebase’s rules protected user identity from admin-level access

Making the app reliable even on weak hostel WiFi

Every challenge made the system stronger, more secure, and more student-centric.

⭐ Accomplishments we’re proud of

Achieved 90%+ accuracy in AI issue classification

Designed a full anonymous complaint ecosystem

Created a transparent tracking cycle used by both students & staff

Fully optimized for Arm architecture (required by the hackathon)

Built a product that can genuinely improve hostel life

⭐ What we learned

We learned how essential anonymity is in environments with strict hierarchies, and how AI + mobile technology can solve simple problems that impact millions of students daily. We also learned how to:

Optimize AI models for Arm-based devices

Build smoother React Native interfaces

Implement secure privacy-first architectures

Design for real-world constraints like low network quality

⭐ What's next for HostelSOS

We plan to add:

Predictive analytics for recurring issues

Community transparency dashboards

Hostel ratings + facility insights

Vendor marketplace for repairs

Gamified engagement for students

Multi-language support (Hindi, Tamil, Telugu, etc.)