sahayya-portal

Sahaya is a multilingual voice-enabled complaint portal where students submit issues in any language. It auto-translates and shows all complaints to admins in their preferred language.

Comment

Inspiration

College complaint systems often fail because students aren’t comfortable expressing issues in English or the local language. Many hesitate to report problems due to language barriers, lack of accessibility, or fear of miscommunication. Sahayya Portal was inspired by the idea of making grievance submission inclusive, voice-enabled, and language-independent so that every student can communicate freely.

Sahayya Portal allows students to submit complaints using voice or text in any language. The system automatically performs:

Speech-to-Text conversion

Language detection

Auto-translation to the admin’s chosen language

Complaint categorization

Organized dashboard view for admins

Admins can view all complaints in one uniform language, while the portal stores both the original and translated versions.

How we built it

We built Sahayya with:

Frontend: React + HTML/CSS

Backend: Node.js

Database: MongoDB

AI APIs: Speech-to-Text + Translation

Architecture: REST-based API with async processing

A pipeline handles:

Audio → STT → Translation → Categorization → Admin Dashboard Audio→STT→Translation→Categorization→Admin Dashboard

Challenges we ran into

Achieving accurate translations for mixed-language (Hinglish/Kannada-English) audio

Handling noisy audio recordings from mobile devices

Designing a clean, frictionless interface for both students and admins

Ensuring real-time translation without lag

Structuring the database to store original + translated complaints efficiently

Accomplishments that we’re proud of

Built a fully functional end-to-end multilingual complaint system

Enabled seamless communication across students and faculty

Created an inclusive platform that supports users from diverse backgrounds

Achieved smooth voice-based interaction with accurate translation

What we learned

Implementing multilingual AI workflows

Understanding how speech recognition varies across accents

Designing accessible systems for real users

Structuring scalable APIs and database models

Improving UX for users who are not tech-savvy

What's next for Sahayya-Portal

Mobile app version with offline voice recording

Analytics dashboard for complaint patterns

Sentiment analysis to prioritize urgent issues

Integration with college authentication systems

Role-based access for departments (hostel, academics, exam cell)

Improving translation accuracy using fine-tuned models

Built With

Share this project:

Updates