Vaultrac — Case Study | Chinmay Laghate

Case Study

Designing Vaultrac for privacy you actually own

A complete product design from concept to investor-ready prototype — a secure, on-device iOS app that turns your digital life into a private, searchable timeline. Zero cloud. Zero compromise.

Role
Lead Product Designer
Scope
iOS App + Brand + Pitch
Duration
16 Weeks
Tools
Figma, Claude AI, Figma Make, Confluence
0
Cloud dependencies — zero bytes ever leave the device
12
Core iOS screens designed across onboarding, timeline & dashboard
3×
Pitch deck formats produced simultaneously with AI
$1.5M
Seed round target — deck delivered investor-ready
01 —

What is Vaultrac?

Vaultrac is a secure, on-device iOS privacy app that silently tracks and organises everything that happens on your mobile — calls, texts, app usage, notifications — into a clean, searchable private timeline. Think of it as your personal black box: a complete record of your digital life that lives entirely on your device and belongs entirely to you.

The business model is freemium SaaS — a free tier for core tracking, Pro at $7.99/mo for unlimited history and AI-powered search, and a Family plan at $14.99/mo for up to five devices. No ads. No data sales. Ever.

iOS App Privacy Tech On-Device AI Consumer SaaS Freemium Figma Make (AI) Design System Investor Pitch

My role

I led end-to-end product design from the initial concept through an interactive Figma prototype and a three-format investor pitch deck. The scope covered product strategy, information architecture, a full iOS component library, all 12 screen designs, brand identity, and a 16-page Confluence documentation set. A key part of the project was using Claude AI as a design thought partner — generating structured content, documentation, and pitch materials at a pace no manual process could match.

The ask

Create a consumer iOS app that captures an individual's entire digital life — on-device, with no cloud dependency — and positions it as the infrastructure layer for the coming wave of personal AI. Then make the business investable: a $1.5M seed deck, financial projections, competitive positioning, and a GTM strategy, all coherent with the product vision.

02 —

What was broken and why it mattered

The problem isn't a product gap — it's a structural injustice baked into the smartphone economy. Three patterns kept surfacing across user interviews and the privacy landscape audit.

01
Billions of moments vanish daily

Every call, text, notification — gone forever with no searchable record. When you need to verify something that happened three weeks ago, it simply doesn't exist. The data was created; you just never got a copy.

02
Your data enriches everyone but you

Big Tech harvests the most intimate signals of your digital life — who you call, what apps you open, when you sleep — to sell to advertisers. You generate the value. You receive targeted ads in return.

03
Zero ownership, zero access, zero recourse

No copy, no control, no benefit. Data brokers profit from your behavioral fingerprint while you have no legal or practical mechanism to access, correct, or delete it from their systems.

"I need to verify a call I had three weeks ago — there's no way to search for it. It just disappeared."

User interview, Discovery phase
03 —

Understanding the landscape and the person

Research split across two tracks: understanding the user's relationship with their own data, and mapping the regulatory and competitive terrain that shapes what's possible on iOS.

Research methods

  • 018 user interviews — privacy-conscious iPhone users ranging from casual to technically sophisticated
  • 02App Store competitive audit — parental monitoring apps, life-loggers, screen time tools, and personal data portals
  • 03Apple platform study — App Tracking Transparency, ScreenTime API, on-device AI, HealthKit access patterns, and App Review guidelines
  • 04Regulatory landscape mapping — GDPR, CCPA, ATT, and data minimisation obligations relevant to a tracking product
  • 05Job-to-be-done mapping across three user archetypes: the Privacy Hawk, the Accountability Seeker, and the Lifelogger

The clearest insight from research: users don't want to think about privacy — they want to feel it. Architecture guarantees mean nothing if the product doesn't surface visible, tangible proof of control.

04 —

Design system first, then AI-directed screens

Rather than designing screens in sequence, I built the token layer and component library first — then used Figma Make to generate layout directions from the design system, reviewing and directing output instead of assembling pixels by hand.

Week 1–3
Discover
User interviews, App Store audit, Apple platform research, regulatory mapping, competitive positioning.
Week 4–6
Define + Design System
Personas, IA, user journeys, job stories — and building the full design system: color tokens, type scale, spacing rules, 30+ iOS component library, and a Figma Make documentation connector.
Week 7–12
Generate + Refine with Figma Make
Used the design system as AI context to generate layout directions with Figma Make. Iterated on prompts, reviewed output, and refined screens across onboarding, timeline, search, and the privacy dashboard.
Week 13–16
Prototype + Investor Materials
Interactive Figma prototype across all surfaces. Investor pitch deck produced in three simultaneous formats — HTML interactive deck, PPTX, and PDF — using Claude AI as content and layout partner.
04b —

How Claude AI changed the way I worked

The traditional design-to-investor-deck pipeline has a long, expensive middle: content strategy, copywriting, visual layout, format production. On this project, that step collapsed. Claude AI was the thought partner for every written artifact — from 16 pages of Confluence documentation to all three pitch deck formats, generated in parallel in a single session.

My role shifted from producing content to directing it: briefing the AI on product strategy, shaping its output to match the investor audience, and iterating on tone, hierarchy, and narrative arc. The creative and strategic work stayed entirely with me; the mechanical production work moved to the AI.

Signal Detail
Before Documentation, pitch content, and design rationale written section by section across multiple sessions. Format production (PPTX, PDF) a separate manual step per deliverable.
After Claude generated structured content from product briefs. All 16 Confluence pages created in a single session. HTML, PPTX, and PDF pitch decks generated in parallel.
Gain 3× more content directions explored within the same timeline. Design and Figma time protected for the work AI can't do: interaction judgment, user testing, and taste.
Role Designer as director — briefing, reviewing, shaping, and making the final call on every output.
05 —

Before and after — three surfaces

Three surfaces drove the design effort. Here is what changed across onboarding, the private timeline, and the privacy dashboard.

Onboarding
Before
  • Users confronted with a wall of permissions upfront, triggering denial by default
  • No explanation of what each permission enables or why it's safe
  • No trust-building before the ask — cold start with no value proof
  • Drop-off in the first 30 seconds for privacy-conscious users
After — V1
  • Progressive onboarding: one permission, one screen, with a concrete value frame before each ask
  • Visual demonstration of what the timeline looks like before asking for access
  • "Nothing leaves your phone" shown as a permanent on-screen signal, not fine print
  • Skip paths for each permission — users build trust at their own pace
Private Timeline
Before
  • Native iOS notifications list — chronological, non-searchable, lost after scroll
  • No unified view across calls, texts, app usage, and notifications
  • No filtering, tagging, or cross-app correlation
  • Events disappear with no permanent record
After — V1
  • Unified timeline with activity type icons, time stamps, and full-text search across all event types
  • Filter by category: calls, texts, apps, notifications — or search across all simultaneously
  • Cross-app correlation surface: see the app you opened three minutes after a call
  • Persistent record — nothing auto-deletes until the user chooses
Privacy Dashboard
Before
  • No visibility into what's being tracked or how granular the data is
  • No per-category on/off controls — all or nothing
  • No data export, deletion, or audit capability
  • Trust asserted by marketing copy, not the product itself
After — V1
  • Single-screen data ownership center: see every category, toggle each on/off independently
  • Storage usage displayed per category with clear local-only indicator
  • One-tap full data export and permanent deletion with confirmation states
  • Visual "data never leaves device" indicator — architecture proof, not copy
Vaultrac iOS App — Prototype V1 ↗
06 —

What was shipped

This was a zero-to-one project — no existing product to compare against. Outcomes are measured against the design goals set at the start of the engagement.

0
Cloud dependencies by architecture
Every design decision — onboarding, timeline, search, export — validated against the on-device constraint. No feature required a server call to function.
16
Confluence documentation pages
Complete product documentation set covering vision, go-to-market, financial model, competitive landscape, team, brand, and principles — all investor-ready.
3
Pitch deck formats produced
HTML interactive deck with keyboard navigation, PPTX for live presentation, and a PDF for async sharing — all consistent in visual language and narrative.
$1.5M
Seed round deck delivered
12-slide investor pitch covering problem, solution, market, business model, competitive moat, GTM, financials, and the ask — in every format a seed investor needs.
07 —

What I took away

1
Privacy must be felt, not just stated

Architecture guarantees mean nothing if users can't see them. The privacy dashboard, the local storage indicator, and the no-cloud signal had to be product elements — not footnotes. Designing for trust means making invisible things visible.

2
AI is a force multiplier, not a shortcut

Using Claude to generate pitch content and documentation freed design time for the work AI can't do: the interaction judgment calls, the user interview synthesis, and the aesthetic decisions that make a product feel right. The value is in the division of labor.

3
On-device constraints breed better design decisions

Without the option of offloading to a server, every feature had to earn its place within iOS sandbox limits. That constraint made the product sharper — nothing was added that didn't justify its local storage and background processing cost.

4
Investor materials are a UX problem

A pitch deck is a product. It has a user (the investor), a job to be done (evaluate whether to give you money), and friction points (unclear narrative, buried metrics, inconsistent visual language). Applying the same design rigor to the deck as to the app made both stronger.

5
Designing for invisibility is the hardest brief

Vaultrac runs silently in the background. The design challenge was making its value surface exactly when the user wants it — and nowhere else. Notification patterns, timeline refresh moments, and search result timing all had to respect the ambient nature of the product.

See the full experience

The Vaultrac iOS prototype covers the full onboarding flow, the private timeline, search, and the privacy dashboard — all in Figma.

Open Prototype