Building Trust in Government Services through Accessible UI Systems.

"Led user research and experience strategy to craft a digital-rights platform recognized as the top solution among IIT teams."

"Our team secured 1st place among competing IITs by developing a user-first solution that bridges technology, accessibility, and impact."

Problem
Frontline health workers spent too much time verifying children using slow face-scanning steps. This created long queues and incomplete nutrition entries.

Solution
We designed an AI-first flow using QR codes and conversational AI to make verification faster and reduce friction.

Impact
Cut verification time by half in our tests, helped workers complete more visits per day, and this idea won the Google APAC challenge round between IIT Kanpur and IIT Hyderabad.

How I got into this problem

When we started working on this, I realised something early: the nutrition flow wasn’t the real issue at all.
The bottleneck was the very first thing workers did — identifying the child.

I kept imagining the scenario like a film scene in my head, because that’s how I think.

A small moment that changed everything

Picture this.

Rita, a frontline worker, is standing outside a house on a sunny afternoon.
A restless child.
The app keeps asking her to “Scan again”.
The mother is waiting.
Three more mothers are already behind her.

She hasn’t even reached the nutrition questions yet.

This was the moment that made the problem real for me.
Design only makes sense when you understand what someone actually goes through in that moment.

What clicked during research

I don’t like dumping research slides, so here’s the story version.

Insight 1: The real delay was before the real work

Workers lost 30 to 40 seconds just verifying identity.
So what?
They rushed through nutrition entry, which is the heart of the job.

Insight 2: Face scanning and real field environments don’t get along

Bright sunlight, moving children, crowded lanes — scanning failed often.
So what?
Workers had to repeat the same step multiple times.

Insight 3: Every extra second felt like pressure

When you have many houses to visit, your patience becomes your workload.
So what?
Workers skipped details to save time, leading to inconsistent data.

This was not a UI problem.
This was an environment and workflow problem.

Reframing the challenge

Instead of asking “How do we make nutrition entry better?”
We asked: “How do we help workers start their task instantly?”

That shift changed everything.

Things we explored (super short)

I like to ideate quickly, throw things on the table, and see what sticks.

• QR tags for instant child identification

• NFC cards for households

• Voice lookup in crowded environments

• Conversational AI for recording nutrition naturally

• Fail-safe alternate flows for when QR isn’t available

• Faster summary screens

We prototyped each one like we were testing film shots — fast, intentional, and focused on what the scene demands.

The final direction

We didn’t overcomplicate it. We went for a flow that matched how workers actually behave.

1. QR-based verification

Small printed QR codes for each child.
One scan.
Instant identification.
No sunlight issues. No fidgety toddlers.

2. Conversational nutrition input

Workers often speak faster than they type.
So the app listens.
“Child had rice, dal, and milk today.”
The system converts this into structured data.

It feels natural, almost like talking to someone.

3. A reliable fallback

If QR is missing, they can simply speak:
“Find Maya’s child born in January.”
Or type a quick name search.
No dead ends.

4. A clean summary

They get a short, clear summary to confirm everything before saving.

Why this mattered

After redesigning the first step of the flow…

• Verification became almost twice as fast

• Workers could complete more visits

• There was less frustration and fewer repeated attempts

• Nutrition entries became more complete

• And yes… the project won the Google APAC round between IITK and IITH, which was a big validation of the thinking

But honestly, the biggest win for me was understanding how field UX in India actually works.
Design happens in sunlight, noise, crowds, and pressure — not in Figma.

What I learned personally

This project taught me something that has stayed with me:

When you are designing for people who work nonstop under real pressure, your job is to remove friction, not add features.

It also taught me the kind of designer I want to be — someone who can mix logic, storytelling, and empathy to explain a design clearly.

And maybe that’s why this case study became one of my favourites.

“We started with a messy problem statement and ended with a clean user flow.”