All work
Consumer AI2025

SatvikScan. Turning ingredient-label doubt into a confident, in-store decision.

AI Dietary Compliance App — SatvikScan lets users scan a packaged food label and instantly check whether it complies with their dietary principles — Swaminarayan, Upvas (fasting), Vegan or Vegetarian. The product has to win at the shelf, in seconds, without making the user feel like they’re using software. Buying decisions happen in moments — clarity, not data, is the actual product.

Role
UX/UI Designer
Duration
4 months
Platform
iOS, Android
satvikscan · screens
01Home

Personal welcome + entry to the scan flow

02Scanner

Live capture with framing guidance

03Scan results

Per-rule verdicts + clear next action

04Scan history

Memory layer that makes repeat decisions faster

The problem

Where the pain was.

  • 01

    Ingredient lists are dense and technical — hard to interpret in the few seconds users actually have to decide.

  • 02

    Buying decisions happen in seconds. Users don’t want to research every ingredient on a packet.

Process

How I worked through it.

01

Defined the in-store moment

Mapped the high-friction decision moment in real shopping environments. The product had to win in seconds, with one hand, mid-aisle — not on a sofa with time to read.

02

Designed Scan → Understand → Decide

A three-beat core flow that mirrors how people actually decide: see the product, get a clear verdict, act on it without re-reading the label. No more, no less.

03

Made AI verdicts explainable

Every result shows the why — which ingredients triggered the verdict, and how it ties back to the user’s chosen dietary rules. Confidence comes from reasoning, not a green tick.

04

Built saved history into the loop

Past scans become a memory layer. Repeat purchases get faster over time, and the app starts to feel like a habit rather than a tool.

Early thinking

Flow sketches.

Before pixels, a notebook. The two flows that anchored every later screen — the high-level system flow on the left, and the user's decision branches on the right.

01Overall flow

Scan → Extract → Analyse → Results → Save History

02User flow

Capture → analysis → ‘happy with the result?’ branch

Outcome

The outcome.

Users move from uncertainty to clarity in seconds. Explainable results made the AI feel credible, and repeat scans started building mindful purchasing habits.

Decision time
minutes of doubt
seconds of clarity
Purchase confidence
hesitation
decided
Repeat behaviour
one-off use
habit-forming
Learnings

What I’d carry into the next one.

  • Trust beats speed alone. Fast results matter, but believable results matter more.

  • Clarity is emotional. Users aren’t just processing data — they’re looking for reassurance.

  • Simplicity requires structure. The cleaner the experience feels, the stronger the system behind it has to be.