Down arrow opens its links.
    IO Creative AgencyIO Creative Agency
    • Services
      • Website developmentBuilt, hosted and cared for: fast sites you own.
      • AI developmentAI features inside the tools you already use.
      • Agentic AIAgents that handle repeat work; you approve.
      • Website reviewA measured review of the site you have.
      • AI coursesOnline, and in person in Virginia.

      Featured

      Website review

      A measured review of your site: speed, search and accessibility, each finding with its evidence.

      How the review works
    • Work
      • Case studiesWhat we learned building them, written up.
      • ExperiencesThe client sites we built and still look after.

      How we show work

      Measured, not claimed

      Every client page shows the live site and one number we measured ourselves.

      See our experiences
    • Courses
      • OnlineLive sessions you join from anywhere.
      • In person in VirginiaSmall groups, hands on, in Northern Virginia.
      • Next cohortDates, seats and what you will build.

      Courses

      AI for small teams

      Practical AI you use at work the next day, online or in a room with us.

      See all courses
    • Website review
    • About

    Try: , , ,

    Switch to dark theme
    Book a callBook a call
    • Services
    • Work
    • Courses
    • Website review
    • About
    Book a callBook a call
    1. Home/
    2. Experiences/
    3. Mobilife

    Insurance and savings app · Mongolia

    Life insurance and savings, built for a phone and a short visit

    Mobilife sells education savings and life insurance entirely inside a telecom's own app. We are designing and building it: the screens, the way a parent moves through them, and the animated scenes that make a long-term promise easy to see. It is in development, and this page shows the working build.

    IO Creative·App design, UX, App development·November 17, 2026·5 min read

    Ran on

    • Next.js
    • React
    • Tailwind CSS
    • TypeScript
    The first-open story, recorded from the working build on 2026-11-17. The art reacts as the story moves from one step to the next.
    On this page· The problem

    On this page

    • The problem
    • How we defined it
    • How we solved it
    • What we applied
    • Design
    • Type: one family that reads Mongolian well
    • Colour: midnight and ice
    • Shape and layout: touch first
    • Motion: scenes that react to the user
    • Where it stands
    • Our vision

    The problem

    Insurance and long-term saving are decisions parents make rarely and carefully. The people we are building this for are roughly 28 to 55, subscribers of a telecom operator in Mongolia, with a mid-range Android or an iPhone. They will open the app for a few minutes, between other things, to protect their family and to start saving for a child's education.

    The usual way to do this is a form, a branch visit, and a lot of small print. A tool that asks parents to trust a promise that pays out in fifteen years has to explain itself in a few taps, in their own language, and make the number feel real.

    The product has to be understood and bought in one short visit, on a phone, by someone who is deciding something serious.

    How we defined it

    The client's developers set a few rules early, and we treated them as the brief. They shaped every screen.

    • No sign-up, no password, no login code. The app opens already signed in, because the telecom's own app has already tied the person to their phone number.
    • One identity check, once. The state's e-ID is used a single time to fetch the person's own details and their consent, and again to confirm a child. After that, returning users just see their data.
    • One way to pay. A single payment method, with a monthly deduction day the parent picks and an initial payment set at the start.
    • One journey pattern. Every product is bought through the same step ring and step list, and each step opens in place.

    From that we set three design goals: a parent should understand what they are buying before they pay, should always know which step they are on, and should be able to leave half-way and continue from the same place.

    How we solved it

    The app is a web mini-app that runs inside the telecom's app, not a store app, so there is nothing to install and no separate account. It is built in Next.js and Tailwind, with a mock of the server in front of it, so every screen could be designed and checked before the backend was ready. The screens are in Mongolian.

    Opening the app for the first time shows a short story of four beats: protect the family, save for the children, pick up where you left off, confirm once. A card on each beat shows the kind of notice the app will send, so the story doubles as a tour.

    The Mobilife home screen on a phone: a greeting, a photograph card that offers to work out what a child's university will cost, and an empty savings state with a start tile.Home
    The first step of the savings journey: a list of three children, one selected, each with an age and a suggested amount to save.Who is it for?
    Home, and the first step of the savings journey: the app lists the children it already knows about. Sample data; recorded 2026-11-17.

    Saving for education is four steps: choose the child, set the goal, set the monthly amount, confirm the contract. The goal step is where the product becomes visible. A tree stands for the child's future, and moving one slider changes the sum the plan aims for. The next step shows a jar that fills as the monthly amount grows.

    The goal and monthly steps in the working build: the picture follows the number the parent sets.
    The goal step: a tree above the amount 77,000,000 tugrik, a slider, a cover card and a full-width Continue button.The goal
    The monthly step: a glass jar of coins above the amount 166,000 tugrik, a Yes and No choice and a switch for the government child allowance.The monthly amount
    Steps two and three. The amounts are sample values.

    The last step is the contract itself, written as a short summary first and the full terms below in folding sections, so the parent reads what matters before reading everything.

    What we applied

    UX

    Show the promise before asking for the payment

    Each amount sits next to a picture that moves with it. A parent drags the goal and sees the plan grow, so the number is something they felt, not just read. Help is one tap away on every step, and a half-finished journey is saved and offered again.

    Design

    A component first system

    There are 120 shared components in the app today (counted on 2026-11-17), and a screen is only ever built from them. A missing piece is first drawn as two or three options on a design board, one is chosen together with the client, and only then does it enter a screen. The board marks every part Approved or In review, so nobody guesses what is final.

    Tech

    Built ahead of the backend, and ready to hand over

    The server is mocked behind a typed contract, so the front end and the real backend can meet later without rework. Every action that creates something is protected from a double tap. The code lives in a repository shared with the client's own developers, code only.

    Design

    Type: one family that reads Mongolian well

    All interface text is set in Inter. We tried the logo's own typeface for headings and body and it did not hold up on screen, and Inter covers the Mongolian letters Ө, Ү and the tugrik sign. The round face stays only for the logo. Money is set large and tabular, so amounts line up.

    Colour: midnight and ice

    The ground is a pale ice grey, the text and primary buttons are midnight blue, and the teal accent carries information. The brand's red stays for the logo and the cover card, so it keeps its meaning.

    • Ice ground#EDF2F6 · Page
    • Card#F8FAFC · Surfaces
    • Midnight#0B1A2C · Text and primary
    • Teal#1F7A99 · Accent
    • Brand red#DE0023 · Logo and attention
    The app's light palette and its brand red, from its colour tokens.

    Shape and layout: touch first

    Everything is for a thumb. Corners follow four fixed roles (chip 8, control 12, card 20, sheet 32), so a card never looks like a button. The bottom of every screen is either the navigation or one full-width button, never both and never a floating action. Success and errors arrive as pop-ups, so the page under them never rearranges.

    Motion: scenes that react to the user

    The animated scenes are real interactive animation, not looping decoration. Each one is driven by a single number from the screen: the cover sum morphs a shield from light to strong, the monthly amount fills the jar, the goal grows the tree. We built a small library of them (14 so far) and show the client two or three variants on the design board before one is chosen. They are drawn to match the app's glass and ice look.

    What we learned

    A scene must fill its box, not guess its height

    In one layout the scenes collapsed to nothing, because a percentage height resolved to zero inside a flexible card. Scenes now fill their box absolutely, so they take the size of the space they are given.

    Where it stands

    Mobilife is in development and in review with the client. The first-open story, home, the education savings journey and the contract step are built and run in the browser on sample data. Several parts are still marked In review on the design board, among them the savings tab for several children, the home dashboard, a quick add-money sheet and a QR code on every child's plan. The client sends rounds of changes, and each one is worked through in the same board-first way.

    We have not launched it, and we claim no result from it. Each part on this page is what the working build does today.

    Our vision

    The product can grow the way families do: a child, then a second child, then relatives who add to a plan by scanning a code, then insurance cover that sits beside the savings. The components, the design board and the mock server mean each of those is an addition, not a rebuild. When the backend goes live, the same screens will run on real data.

    On this page

    • The problem
    • How we defined it
    • How we solved it
    • What we applied
    • Design
    • Type: one family that reads Mongolian well
    • Colour: midnight and ice
    • Shape and layout: touch first
    • Motion: scenes that react to the user
    • Where it stands
    • Our vision
    Previous caseIO CreativeWeb design and AI for small business, VirginiaNext caseIO InstituteProgramming school, Mongolia

    The same care, for your site.

    One team builds it, measures it and keeps it running.

    We look at what you have.

    The site, its speed, its search.

    We fix what matters most.

    Measured before and after.

    We keep it that way.

    Hosting, updates and a weekly report.

    A site like this, for you?

    Show us what you run today. We reply with what we would fix first.

    IO Creative AgencyIO Creative Agency

    Websites, AI development and courses, built in Virginia.

    Get the next cohort date

    One email when the next AI course date is set. Nothing else.

    Notify me

    Contact

    hi@iocreative.org

    Services

    • Website development
    • Hosting and care
    • AI development
    • Agentic AI
    • Website review
    • AI courses

    Work

    • Case studies
    • Blog
    • Experiences

    Company

    • About
    • Contact
    • Book a call

    © 2026 IO Creative Corporation

    • Privacy
    • Terms
    • Accessibility