If you've ever tried searching for government welfare schemes, you probably know how overwhelming it can get.
Different websites. Complicated eligibility criteria. Language barriers. Endless searching. And the constant question:
"Am I even eligible for this?"
I built Sahaaya Setu to tackle exactly this problem.
"Sahaaya" means support, and "Setu" means bridge. The idea was simple: create a bridge between citizens and the benefits they deserve.
What does Sahaaya Setu do?
Instead of making users manually search through multiple portals, the platform asks for a few details such as:
- Name
- Age
- Gender
- Educational qualification
- State
- Caste category
- A short self-description
Based on these inputs, the system matches users with potentially relevant schemes.
But it doesn't stop there.
For every recommended scheme, users receive:
- A match scorecard
- A simple explanation of why they qualify
- The documents required
- A direct link to the official government portal
- Step-by-step guidance to proceed with the application process
The goal wasn't to replace government websites. The goal was to make the journey towards them easier.
Multilingual by Design
India is incredibly diverse, and accessibility should include language accessibility too.
Sahaaya Setu was designed with support for 10+ Indian regional languages, allowing users to interact with the platform in the language they are most comfortable with.
Information becomes more useful when people can actually understand it.
The AI Layer
One feature I was particularly excited about was the AI assistant integrated into the experience.
The AI bot functioned as a personalized guide that could:
- Answer frequently asked questions,
- Explain schemes in simpler terms,
- Clarify doubts,
- Provide guidance throughout the process.
Instead of overwhelming users with legal or administrative language, the aim was to create an assistant that felt approachable and easy to understand.
Building the Project
Sahaaya Setu was built as part of a hackathon organized by Be10X.
The project was developed using Replit and Perplexity, which allowed me to rapidly prototype and experiment with ideas.
Hackathons teach you to move quickly, prioritize features, and constantly adapt as new challenges arise.
Mistakes I Made
No project is perfect, and this one certainly wasn't.
One limitation was that the AI assistant itself wasn't multilingual, even though the broader platform supported multiple languages. If I were to improve the project, expanding multilingual support to the AI layer would be a priority.
I also discovered an issue with the eligibility logic for the National Scholarship Scheme. The scheme has age-related eligibility requirements, but my implementation recommended it to users across all age groups instead of properly enforcing the age criteria.
It was a valuable reminder that when building systems that provide guidance, accuracy matters just as much as innovation.
What I Learned
Building Sahaaya Setu taught me that technology doesn't always have to reinvent the wheel.
Sometimes, its greatest strength lies in making existing systems easier to navigate.
A small improvement in clarity can save someone hours of confusion.
Accessibility is not only about whether information exists.
It's also about whether people can understand it, trust it, and act upon it confidently.
Looking Ahead
Sahaaya Setu began as a hackathon project, but the problem it addresses is very real.
If I continue working on it, I would focus on:
- Improving eligibility validation,
- Expanding multilingual AI support,
- Enhancing personalization,
- And ensuring the recommendations remain accurate and transparent.
At its heart, Sahaaya Setu represents something I deeply care about:
using technology to bridge gaps between people and opportunities.
And if even one person can move from days of searching to minutes of clarity, then I think that's a meaningful place to begin.
Comments
Post a Comment