Saathi was an AI-first experiment — how much could AI compress a full design process into one week?
Research, IA, and screens were all AI-assisted. My role was directing and deciding, not delegating.
Role
End to End Product Designer
Timeline
2026 | 1 week
Project
Experiment
Tools
Problem Statement
A prescription shouldn't require a translator.
In India, prescriptions are often handwritten in English and filled with medical abbreviations, making them difficult for many patients to understand.
This creates challenges such as:
Missed doses
Incorrect medication usage
Dependency on family members
Fear of asking healthcare professionals for clarification
I explored existing studies, healthcare reports, and accessibility challenges to understand the broader context before talking to users.
Activities
Literature review
Competitor exploration
Healthcare workflow analysis
Accessibility benchmarking
73%
Patients struggle to fully understand prescriptions
52%
Medication non-adherence rate
125k
Deaths annually linked to medication misuse
22
Official Indian languages, while most prescriptions remain in English
Defining Problem Statement
In India, a prescription is written in seconds and handed over in silence. For millions of patients elderly, low-literacy, non-English speaking, that piece of paper is unreadable. Missed doses and wrong usage follow. The system wasn't designed to harm. It just never learned to communicate. Saathi asks: what if the prescription could explain itself?
Research
The Stories Behind The Statistics
I sat with 6 elderly patients and 3 pharmacists & doctors across Maharashtra, asking each the same set of questions. I shadowed pharmacy visits and collected real prescriptions, the same confusion surfaced in every conversation.
Sunita Joshi
58 • Homemaker • Nashik
"I need someone to tell me, in Marathi, exactly when and how to take my BP medicines without making me feel ashamed."
Pain Points:
English Rx
Doctor speaks fast
Forgets evening dose
Rajesh Kumar
62 • Retired clerk • Mumbai
"I have 4 medicines for diabetes & heart. I want a clear schedule that won't let me skip but I don't want to look like a confused old man."
Pain Points:
Too many similar pills
Side effects scare him
Insurance paperwork
Priya Kumar
34 • Daughter • Bengaluru
"I live 1,500 km from my parents. I need to know they took their pills today without calling them at 6 AM."
Pain Points:
Worry from distance
Parents under-report misses
Confusing pharmacy receipts
Meera Kapoor
41 • Pharmacist • Mumbai
"I want a 60-second tool that explains dosing visually so I can serve more patients without skipping counseling."
Pain Points:
Queue pressure
Repeating same explanations
Liability worries
Key Insight
Patients weren't failing healthcare.
Healthcare was failing communication.
Opportunity mapping
Where intervention has the most leverage
I mapped every pain point to an opportunity, then scored each on impact (patient outcome) vs effort (engineering + clinical). The top 9 became my MVP.
Translate handwritten Rx → pictograms
Voice playback per medicine
Visual dosage timeline
Family caregiver loop
Pharmacist-assist video
Side-effect emergency triage
Offline reminders
Refill predictions
Subsidy & Jan Aushadhi nudges
Information architecture
5 top-level destinations. Nothing more
For low-literacy users, every extra menu is a barrier. The IA collapses the entire app into 5 tabs, with the scan button as a persistent center action, the most-used flow is always one tap away.
Competitive landscape
Where Current Solutions Fall Short
Existing apps optimize for urban smartphone-native users. Pictogram support, regional languages, and pharmacist assist are rare or absent.
Design system
Warmth, clarity, dignity.
A system designed for elderly eyes, regional scripts, and ₹6,000 phones. Built in pixels you can see from arm's length, colors that survive sunlight, and pictograms that work in any language.
The Final Experience
From Confusion to Confidence
Logo
Onboarding
Language selection
Scan Prescription
Medicine Details
Home
Voice Player
The final screens are only half the story. Explore the research, insights, iterations, and decisions that shaped Saathi.
Saathi — an AI-first one-week experiment converting handwritten prescriptions into pictograms, voice playback, and reminders to improve adherence and understanding.