Resume

EDTECH

B2B PRODUCT

Preventing Student Drop-Out with

Proactive Risk Analysis

Designing a teacher dashboard that transforms passive student data into real-time, actionable risk insights before it's too late.

PLATFORM

Web and mobile app

DURATION

12 Weeks

MY ROLE

Product designer

WHAT WE ACHIEVED

36k +

Daily active users using across 200 + institutes

20%

Reduction in student churn rate MOM

2 Cr.

Annual saving in tech.

SHOWREEL

THE PROBLEM

Students disappear silently.
Teachers find out too late.

In India's competitive coaching institutes, 30–40% of enrolled students disengage months before they formally drop out. Attendance is tracked in spreadsheets. Risk is felt, not measured. Teachers have no system to spot patterns and by the time they notice, the student is already gone.

No Early Warning System

Attendance data lived in siloed spreadsheets. There was no automated flag for declining engagement patterns. Teachers relied entirely on memory.

Reactive, Not Proactive

Follow-ups happened after the drop-out, not before. By the time admin noticed dipping ratings or batch absences, the student had already mentally checked out.

Desktop-Only Tools for Mobile Teachers

Follow-ups happened after the drop-out, not before. By the time admin noticed dipping ratings or batch absences, the student had already mentally checked out.

UNDERSTANDING THE USERS

Two roles, one shared goal

Research revealed two distinct Audience in the ecosystem, each with different levels of data access, urgency, and context switching needs.

Priya - Senior Faculty Member

5 batches • 100+ students

"I know Ankit has been absent but I don't know if it's serious or just a cold. I need context, not just a number."

• Needs a quick risk summary before entering class

• Wants to take action from her phone between sessions

• Needs historical context, not just today's attendance

• Doesn't have time to navigate complex dashboards

Rahul — The Batch Admin

Manages 500+ students across batches

"Every month I'm surprised by dropouts. I need to see where the risk is building before the parents call me."

Needs an at-a-glance overview of all at-risk students
Wants to assign follow-up tasks to teachers
Needs to track if follow-ups were actually done
Requires export-ready data for parent meetings

DESIGN PROCESS

From signals to system

A 4-phase process that prioritised talking to teachers before touching Figma, and shipping fast iterations over perfect designs.

Discovery & Research

Conducted 15+ contextual interviews with teachers across 8 institutes.
Shadowed an admin for half a day to observe how attendance was actually being logged.

Key finding: risk was defined intuitively but never measured. This was the design opportunity.

Contextual Inquiry

15+ interviews

Competitive Audit

Affinity Map

Information Architecture & Flows

Categorised content according to relevance . Mapped out the core journey to review
at risk students and take actions. Tested the IA with card sorting across 4 participants.

User Flow Mapping

Card Sorting

Content Priortization

Iterative Design & Usability Testing

Created MVP designs for stakeholder alignment. Iterated designs further more to make
sure upmost important information is flagged to faculties on dashboard. Designs
for AI summary feature. Ran 3 rounds of usability testing with 5+faculties.

Lo-Fi Wireframes

MVP Iterations

2+ Usability Test Rounds

Handoff & Validation

Delivered a dual surface design system (desktop + Mobile). Annotated all edge
cases including empty states, loading states, error scenarios, and offline mode for mobile.

Dev Handoff

Edge Case Documentation

UAT

THE SOLUTION

One dashboard. Two surfaces.
Zero guesswork.

The dashboard surfaces the right signal to the right person giving teachers a fast mobile view and admins a rich desktop overview, all anchored to the same live risk data.

STUDENT PROFILE

All student insights visible
at a glance

All faculties and admin have access to all of student attendance, mark sheets at a glance. faculties can look at LOP of students. The focus shifts from taking actions at the last moment to taking action at the begining.

AI INSIGHTS

Leveraging AI to
reduce churn

AI scans through all student profile and suggests students with hight probability of churn to the faculties and admin. Faculties can furthermore consolidate student profiles and know exactly what issue the students are facing.

IMPACT

What Changed After Launch

Measured across the pilot cohort of all institutes over the first 8 weeks post-launch.

3X

More Faculty - student interactions

-80%

Reduction in follow up TAT

90%

Faculties onboarded within first 8 weeks

LEARNINGS

Reflections & Next Steps

Data dashboards need a "so what" layer

Showing 72% risk score means nothing without telling the teacher what to do next.
The best moment in the project was when we added the Follow Up panel — suddenly data had a destination. Any B2B dashboard without a clear next action is just noise.

Teachers are human, not admins

Early designs were cold and clinical. Adding the avatar illustration on the dashboard ("Hi Dhruv") and humanising student names in the follow-up panel made the product feel like a collaborator, not just a tool. Emotional resonance drives adoption even in enterprise products.

Predictive nudges + parent communication layer

The next version should push proactive notifications to teachers before a student hits the Moderate threshold — not after. A parent communication feature is also on the roadmap: allowing teachers to trigger templated messages directly from the student detail screen.

Innovaccer Analytics (WIP)

Creating agentic workflow for care managers

What’s New

Pepperoni Makhni Pizza

Taco Maxicana

Tap to stop video

Delicious symphony of pepperoni, onions and melting cheese that will fill your mouth with flavour and leave your craving for more

₹320

Add

🎉 For your special events

View all

Birthday

Parties ?

Order Now

Hungry at parties?

Order Now

Payment Offers

View all

₹80 Cashback

Upto ₹200 on order of min. ₹300 on Hdfc credit & debit cards

More details

Apply

₹80 Cashback

Upto ₹200 on order of min. ₹300 on Hdfc credit & debit cards

More details

Apply

₹80 Cashback

Upto ₹200 on order of min. ₹300 on Hdfc credit & debit cards

More details

Apply

Still Searching ?

Veg Pizza

Non-Veg Pizza

Taco

Pasta

Pepsi

Sides

Garlic Bread

Choco Lava Cake

Explore menu

తెలుగు

ਪੰਜਾਬੀ

Introducing Virtual host

Select Language

Virtual host helps you to understand the items in your native language now.

Skip

What’s New

Pepperoni Makhni Pizza

Taco Maxicana

Tap to stop video

Delicious symphony of pepperoni, onions and melting cheese that will fill your mouth with flavour and leave your craving for more

₹320

Add

🎉 For your special events

View all

Birthday

Parties ?

Order Now

Hungry at parties?

Order Now

Payment Offers

View all

₹80 Cashback

Upto ₹200 on order of min. ₹300 on Hdfc credit & debit cards

More details

Apply

₹80 Cashback

Upto ₹200 on order of min. ₹300 on Hdfc credit & debit cards

More details

Apply

₹80 Cashback

Upto ₹200 on order of min. ₹300 on Hdfc credit & debit cards

More details

Apply

Still Searching ?

Veg Pizza

Non-Veg Pizza

Taco

Pasta

Pepsi

Sides

Garlic Bread

Choco Lava Cake

Explore menu

Introducing famous near you

We are introducing the locally famous items around you. Now enjoy the best sellers in your area!

The Locally Famous

View All

Delicious vegetarian pizza topped with fresh local ingredients like mushroom, bell pepper, onions

Farmhouse Pizza

Delicious vegetarian pizza topped with fresh local ingredients like...Read More

₹320

Regular | Cheese Burst

Add

Layered taco filled with spicy tandoori sauce and vegetable pattie

Taco Maxicana

Delicious taco with creamy tandoori sauce and vegetable patties to fill.. Read More

₹320

Add

Skip

Dominos India

Creating agentic workflow for care managers

Preventing Student Drop-Out with

Proactive Risk Analysis

Designing a teacher dashboard that transforms passive student data into real-time, actionable risk insights before it's too late.

Web and mobile app

12 Weeks

Product designer

36k +

36k +

Daily active users using across 200 + institutes

20%

20%

Reduction in student churn rate MOM

2 Cr.

Annual saving in tech.

Students disappear silently.Teachers find out too late.

In India's competitive coaching institutes, 30–40% of enrolled students disengage months before they formally drop out. Attendance is tracked in spreadsheets. Risk is felt, not measured. Teachers have no system to spot patterns and by the time they notice, the student is already gone.

No Early Warning System

Attendance data lived in siloed spreadsheets. There was no automated flag for declining engagement patterns. Teachers relied entirely on memory.

Reactive, Not Proactive

Follow-ups happened after the drop-out, not before. By the time admin noticed dipping ratings or batch absences, the student had already mentally checked out.

Desktop-Only Tools for Mobile Teachers

Follow-ups happened after the drop-out, not before. By the time admin noticed dipping ratings or batch absences, the student had already mentally checked out.

Two roles, one shared goal

Research revealed two distinct Audience in the ecosystem, each with different levels of data access, urgency, and context switching needs.

Priya - Senior Faculty Member

5 batches • 100+ students

"I know Ankit has been absent but I don't know if it's serious or just a cold. I need context, not just a number."

• Needs a quick risk summary before entering class

• Wants to take action from her phone between sessions

• Needs historical context, not just today's attendance

• Doesn't have time to navigate complex dashboards

Rahul — The Batch Admin

Manages 500+ students across batches

"Every month I'm surprised by dropouts. I need to see where the risk is building before the parents call me."

Needs an at-a-glance overview of all at-risk students

Wants to assign follow-up tasks to teachers

Needs to track if follow-ups were actually done

Requires export-ready data for parent meetings

From signals to system

A 4-phase process that prioritised talking to teachers before touching Figma, and shipping fast iterations over perfect designs.

Discovery & Research

Conducted 15+ contextual interviews with teachers across 8 institutes. Shadowed an admin for half a day to observe how attendance was actually being logged. Key finding: risk was defined intuitively but never measured. This was the design opportunity.

Information Architecture & Flows

Categorised content according to relevance . Mapped out the core journey to review at risk students and take actions. Tested the IA with card sorting across 4 participants.

Iterative Design & Usability Testing

Created MVP designs for stakeholder alignment. Iterated designs further more to make sure upmost important information is flagged to faculties on dashboard. Designs for AI summary feature. Ran 3 rounds of usability testing with 5+faculties.

Handoff & Validation

Delivered a dual surface design system (desktop + Mobile). Annotated all edge cases including empty states, loading states, error scenarios, and offline mode for mobile.

One dashboard. Two surfaces.Zero guesswork.

The dashboard surfaces the right signal to the right person giving teachers a fast mobile view and admins a rich desktop overview, all anchored to the same live risk data.

All student insights visibleat a glance

All faculties and admin have access to all of student attendance, mark sheets at a glance. faculties can look at LOP of students. The focus shifts from taking actions at the last moment to taking action at the begining.

Leveraging AI to reduce churn

AI scans through all student profile and suggests students with hight probability of churn to the faculties and admin. Faculties can furthermore consolidate student profiles and know exactly what issue the students are facing.

What Changed After Launch

3X

3X

More Faculty - student interactions

-80%

-80%

Reduction in follow up TAT

90%

90%

Faculties onboarded within first 8 weeks

Reflections & Next Steps

Data dashboards need a "so what" layer

Showing 72% risk score means nothing without telling the teacher what to do next. The best moment in the project was when we added the Follow Up panel — suddenly data had a destination. Any B2B dashboard without a clear next action is just noise.

Teachers are human, not admins

Early designs were cold and clinical. Adding the avatar illustration on the dashboard ("Hi Dhruv") and humanising student names in the follow-up panel made the product feel like a collaborator, not just a tool. Emotional resonance drives adoption even in enterprise products.

Predictive nudges + parent communication layer

The next version should push proactive notifications to teachers before a student hits the Moderate threshold — not after. A parent communication feature is also on the roadmap: allowing teachers to trigger templated messages directly from the student detail screen.

Students disappear silently.
Teachers find out too late.

Conducted 15+ contextual interviews with teachers across 8 institutes.
Shadowed an admin for half a day to observe how attendance was actually being logged.

Key finding: risk was defined intuitively but never measured. This was the design opportunity.

Categorised content according to relevance . Mapped out the core journey to review
at risk students and take actions. Tested the IA with card sorting across 4 participants.

Created MVP designs for stakeholder alignment. Iterated designs further more to make
sure upmost important information is flagged to faculties on dashboard. Designs
for AI summary feature. Ran 3 rounds of usability testing with 5+faculties.

Delivered a dual surface design system (desktop + Mobile). Annotated all edge
cases including empty states, loading states, error scenarios, and offline mode for mobile.

One dashboard. Two surfaces.
Zero guesswork.

All student insights visible
at a glance

Leveraging AI to
reduce churn

Showing 72% risk score means nothing without telling the teacher what to do next.
The best moment in the project was when we added the Follow Up panel — suddenly data had a destination. Any B2B dashboard without a clear next action is just noise.