AWS Builder Center
GovEase - Day 2: From Foundation to a Live AI-Powered Product šŸ‡®šŸ‡³

GovEase - Day 2: From Foundation to a Live AI-Powered Product šŸ‡®šŸ‡³

Day 2 of the First Commit Hackathon — GovEase is LIVE! šŸš€

Student Fellow @Databricks | AWS Certified (x5) | Skill Builder | Multi-Cloud & AI Pro | Microsoft |
First Commit Hackathon | Bharat Builds Tour | WeMakeDevs Ɨ AWS
Yesterday, GovEase was a dataset of 47 government schemes stored in DynamoDB.
Today, it became a working product.
The goal is simple:
Answer 7 questions → discover government schemes that match your profile.
No signup. No unnecessary complexity. AI-powered recommendations.



🧠 Building the Eligibility Engine

The biggest task for Day 2 was turning the scheme data into actual application logic.
I built an AWS Lambda function using Python that:
  • Reads the 47 schemes from DynamoDB
  • Checks age, income, gender, category, state and occupation
  • Calculates eligibility matches
  • Scores and ranks the results
  • Sends the matched schemes to Amazon Bedrock
The backend now turns a user's profile into a ranked list of potentially relevant schemes.



šŸ¤– Adding Amazon Bedrock

Once the eligibility engine identifies matching schemes, the results are passed to Amazon Bedrock using Nova Lite.
The purpose isn't simply to generate another list.
The AI creates a personalized explanation of the matched schemes and helps the user understand which opportunities may be worth looking at first.

First Test

I tested GovEase with a profile representing a:
20-year-old OBC student from Tamil Nadu
The system returned:
18 matching schemes in 2.6 seconds.
That was the first moment when the project felt like more than a database.
The data was now becoming an actual user experience.



šŸ”Œ Connecting Everything With API Gateway

Next, I connected the Lambda backend to the frontend using Amazon API Gateway.
The API uses:
  • REST API
  • POST endpoint
  • Lambda proxy integration
  • CORS enabled
The complete request flow is now:
User Profile
↓
Frontend
↓
API Gateway
↓
AWS Lambda
↓
DynamoDB
↓
Eligibility Matching
↓
Amazon Bedrock
↓
Personalized Results




šŸŽØ Building the Frontend

For the frontend, I wanted something lightweight and simple.
Instead of introducing a large framework, I built the initial interface using a single HTML file.
No React. No npm. No build step.
The visual identity uses India's tricolor as inspiration:
  • Saffron for actions and accents
  • Green for benefits and success states
  • Navy blue for headings and visual elements
The focus was to keep the experience simple:
Enter your details → Check eligibility → Understand your options.



🌐 Going Live

By the end of Day 2, the complete flow was connected and deployed through Amazon S3.

Architecture

Amazon S3
↓
API Gateway
↓
AWS Lambda
↓
DynamoDB
+
Amazon Bedrock




šŸ’° Keeping It Serverless

One of the reasons I chose this architecture was cost and simplicity.
Using managed AWS services means I don't need to maintain a traditional server for the application.
For this hackathon build, the AWS usage was effectively ₹0 using available free-tier and hackathon resources.



šŸŽÆ What Day 2 Taught Me

Day 1 was about building the foundation.
Day 2 was about connecting the pieces.
I learned that getting individual AWS services working is only the beginning.
The real challenge is making them work together:
Database → Backend → AI → API → Frontend → Deployment
And today, that complete flow finally worked.



šŸš€ What's Next — Day 3

The product is live.
Now comes the polishing phase:
  • Test different user profiles
  • Handle edge cases
  • Fix bugs
  • Improve the UI
  • Record the demo
  • Prepare the final hackathon submission
Day 1 gave GovEase a foundation. Day 2 turned it into a product. Day 3 is about getting it ready to present. šŸ‡®šŸ‡³šŸš€
Any opinions in this article are those of the individual author and may not reflect the opinions of AWS.
Enjoyed reading this content? Let the author know!

Your likes, comments, shares, and saves help creators reach more builders.

Loading recommendations

Loading article