DramaRemix

Watch a story.
Make it your own.

An AI-powered short-drama app that lets everyday viewers create new storylines from the stories and characters they love.

  • AI-Powered
  • 0→1 UI Design
  • Interaction Design
  • Usability Testing
  • UX Prioritization
  • Design System
Role
UX / UI Design Intern
Timeline
June – September 2026
Duration
4 months
Platform
iOS
Two DramaRemix playback screens in iPhone mockups

Overview

What is Drama Remix

DramaRemix connects watching and creating in one experience. Viewers can explore story branches and use AI to create their own short dramas with familiar characters.

Team & my role

Designer · me
1
Product manager
1
Engineers
3

As the team’s UI / UX Design Intern, I worked with a product manager and three engineers across two stages:

Building the UI foundation. Our first MVP aimed to connect watching short dramas, creating new storylines, and returning to created work in one usable experience. I helped turn that goal into page layouts, reusable components, and interface states, giving the team a consistent UI foundation for the initial release.

Refining key experiences. I independently audited the creation flow and Personal Center, then used usability findings to redesign creation steps, clarify generation feedback, and improve how users find and manage their work.

Timeline
JuneJulyAugustSeptember
◆ Launch
  1. Design audit
  2. User research · interviews
  3. UI foundation · v1
  4. Creation Center
  5. Personal Center
  6. Topics & collections
  • Before launch
  • After launch

My contribution

0 → 1

Built the UI foundation

Turned product requirements into page layouts, reusable components, and interface states.

Outcome

Screens Designed
60+
Design Iterations
2
Usability-Test Participants
15

Design Audit

To explore this new viewing-to-creation experience, we relied primarily on testing with real users to validate how it worked for them.

Walk through the existing screens and key tasks: watching, choosing a branch, and creating a story.

Capture unclear interactions, missing feedback, and inconsistent UI patterns to validate with users.

The design audit and moderated testing with 15 participants surfaced 12 issues. These three themes guided the design.

Creation Center

Unclear next step

Major · Task-blocking

Story context and controls compete for attention.

Creation Center

Unclear AI feedback

Major · Task-blocking

Progress, credit use, and refresh behavior are unclear.

Personal Center

Hard to find created work

Major · Task-blocking

Users cannot tell where their published branches appear.

Explore the other 9 observations

Story choices

Time to understand

Major · Task-blocking

The countdown adds pressure before users can understand and compare the options.

Story choices

Cost & destination

Major · Task-blocking

Similar cards obscure what each action does, what it costs, and what users receive.

Story tree

Where am I?

Major · Task-blocking

The current story position and the meaning of solid and dashed lines are not clear.

Story tree

Which branch should I open?

Major · Task-blocking

Cards provide limited context for comparing branches or identifying who created them.

Personal Center

Labels need a clear meaning

Major · Task-blocking

Terms such as “My Theater” and “Activity” do not clearly communicate what they contain.

Playback & entry

Mode awareness

Major · Task-blocking

The current mode and the effect of switching are unclear.

Playback & entry

Interaction entry

Major · Task-blocking

The game-like icon obscures story interactions, and a direct story-tree entry is missing.

Visual consistency

Tag styling

Minor · Non-blocking

Standardize tag colors and shapes across screens.

Visual consistency

Repeated patterns

Minor · Non-blocking

Align view-count badges and recurring UI styles.

Phase 1 established a consistent UI foundation across typography, tags, and shared patterns. Phase 2 framed and refined the three core flow blockers.

Working hypothesis · Relative impact on task completion and design-side effort estimates, pending engineering validation.

Prioritization strategy: impact versus effortImpact on task completion increases upwards; effort increases to the right. Coral orange highlights the three core flow blockers: next-step clarity, AI feedback, and access to creations. Golden orange highlights UI consistency addressed in Phase 1. Coral orange marks Phase 2 core flow blockers. Gray indicates other addressed insights, not unresolved work. Lower impact means lower impact on task completion. All placements are provisional.High ImpactLowHigh Effort01 · Competing controlsobscure the next step02 · Uncertain AI statusundermines confidence03 · Hidden creationsbreak the return journey04 · Countdown pressurelimits informed choices05 · Unclear costsmake choices uncertain06 · Ambiguous branchesmake exploration difficult07 · Hidden interaction routesinterrupt discovery08 · Inconsistent tagsslow visual scanning09 · Inconsistent UI patternsweaken recognition

05UX Problem Framing

Core Pain Points

01
Unclear next step

Story context and controls compete for attention.

02
Unclear AI feedback

Progress, credit use, and refresh behavior are unclear.

03
Hard-to-find creations

Viewing-history tabs obscure access to created work.

How Might We

How might we make the next creative step obvious?

How might we clarify what the AI is doing and what comes next?

How might we help creators recognize and return to their work?

Design Opportunities

Guide one creative decision at a time.

Make progress, cost, and recovery clear.

Make creations easy to recognize and find.

Design Strategy

Phase 1

Build the UI foundation Turn requirements into a consistent app.

  1. Foundations

    Define visual foundations

    Unify color, typography, icons, and corner styles.

  2. Components

    Build reusable components

    Standardize buttons, navigation, loading, and interface states.

  3. Application

    Apply across the app

    Translate requirements into page layouts using the shared system.

Phase 2

Refine the core experience Help creators move forward with confidence.

  1. Creation flow

    Guide the creation flow

    Separate character selection, story direction, and final review.

  2. AI feedback

    Clarify AI feedback

    Make generation status, credit cost, and recovery actions explicit.

  3. Personal Center

    Bring creations forward

    Prioritize created work and its performance in the Personal Center.

Phase 1 · UI Foundation

Design System

Color roles
18
Specification boards
7
Icon assets & variants
75
Text styles
18

Hover to exploreSwipe to explore

DramaRemix brand identity across sizes and formats.
Eighteen text styles with sizes, weights, usage, and line heights.
Eighteen color roles for brand, actions, text, surfaces, points, and feedback.
Seventy-five icon assets and variants for interaction, navigation, and playback.
Buttons, navigation, playback progress, action bars, and credit indicators.
Corner radius specifications with original interface examples.
Network errors, permissions, loading, placeholders, and confirmations.

Grid & Spacing

A consistent grid and spacing system across screens.

Content grid
4 columns
Screen width
430 px
Side margins
12 px
Column gutter
8 px
Grid size
4 px
4 column Grid8px8px8px
DramaRemix voting screen exported with its Figma layout grids: four columns, 8 px gutters, 12 px side margins, and a 4 px grid

UI Design

Page layouts, reusable components, and interface states for the product’s first complete experience.

With the UI foundation in place, Phase 2 focused on creation and helping users find their work.

Phase 2 · UX Redesign

Iteration 1: Creation Flow

Before

Original Creation Center with callouts identifying unclear context, selection hierarchy, step order, relationship labels, and entry points

“Reference direction” lacksclear story context.

Primary and secondary choiceshave no clear hierarchy.

Creative choices haveno clear step order.

Relationship labels do notidentify the characters.

Typing and tag selectioncompete as the first action.

Redesigned Creation Flow

From the existing story to a new creationChoose characters, set relationships, choose a story direction, optionally add a title and description, review the prompt, and start generation. The optional step can be skipped.Skip optional details01Choose charactersFrom the existing story02Set relationshipsBetween the characters03Choose a directionShape the next storyline04Title & descriptionOptional05Review the promptCheck before creating06Start generationCreate the next story
  1. Character selection

    Choose the characters

    A clear starting point. Choose familiar characters from the original story. Keeping one section open helps users focus on the current choice.

    Choose the characters: Character selection in the redesigned Creation Center
  2. Story direction

    Shape the story

    One creative decision at a time. Choose a story direction in its own section, while earlier choices remain visible as compact summaries.

    Shape the story: Story direction in the redesigned Creation Center
  3. Prompt review and credit cost

    Review before creating

    Confirm before committing. Review the assembled prompt and see the credit cost before generating the story.

    Review before creating: Prompt review and credit cost in the redesigned Creation Center

Before & After

A clear starting point, one decision at a time, and a visible next action.

Before

Original Creation Center with competing controls

After

Redesigned Creation Center with a source story, separate character selection, numbered sections, optional details, and a Next button

The source story establishesa clear starting context.

Character choices have theirown section and selection limit.

Numbered sections makethe decision order explicit.

Relationship choices are separatefrom the story direction.

A clear Next action advancesone focused step at a time.

Outcome

Creation flow · Selection to prompt review

Users can create a new storyline in less than two minutes with the redesigned flow.

Before

Viewing History / Creator CenterWatching comes first,but creators need their work.

History / Favorites / LikesFavorites is buried beneathless useful viewing history.

Episode progress / Last viewedCards show viewing progress,not how creations perform.

Redesigned Personal Center Sitemap

Information hierarchy within Profile. Connections show where content belongs, rather than separate pages or sequential steps.

  • Profile
    • Wallet
      • Coin Balance
    • Profile Information
      • Share Profile
      • Avatar
      • User ID
      • Username
      • Edit Profile
    • Social Connections
      • Followers
      • Following
    • Promotional Banner
      • Invite Friends
      • Tasks
    • Creator Center
      • My Creations
        • Tasks
        • Story Branches
          • Creation Analytics
            • Branch Count
            • Total Views
            • Creation Rewards (Coins Earned)
    • Favorites
      • All
      • Series
      • Story Branches
      • + New Folder
    • Viewing History
      • Series
      • Story Branches

Before & After

After
Before
After

Keep your profile connected

Followers, Following and Edit Profile stay together beside your identity.

Improvement 1 of 4

Final Design

Playback & Story Branching

Watch — final product interface

Watch

Watch a story, then explore another direction.

Choose — final product interface

Choose

Explore the available branches at a story moment.

Explore — final product interface

Explore

See how the story branches connect.

Story Creation Flow

Cast — final product interface

Cast

Start with characters from the original story.

Direct — final product interface

Direct

Choose where the story goes next.

Review — final product interface

Review

Review the prompt before generating a video.

Personal Center & Discovery

Your work — final product interface

Your work

Find creations and their performance in one place.

Discover — final product interface

Discover

Browse stories and continue watching.

Topics — final product interface

Topics

Explore stories around shared interests.

Selected final product interfaces. The story tree provides product context, outside the iteration work discussed above.

Reflection

  1. 01

    Design for the whole experience

    I learned to consider how watching, creating, and returning to previous work connect, and to evaluate individual design decisions within that broader journey.

  2. 02

    Prioritize information around user goals

    I learned to prioritize information by its relevance to the user’s current task—from making story choices to checking video performance—and keep secondary details available without competing for attention.

  3. 03

    Use AI with a stronger design foundation

    Working with AI across initial prototypes, usability-test analysis, prioritization, revised prototypes, and design-system implementation showed me both its strengths and its limits. In the product, users needed clear creative choices and useful feedback—not a walkthrough of the AI’s internal process. In my own workflow, generating interfaces without sufficiently defined rules led to inconsistent components and unbalanced layouts that took considerable effort to review and correct. A solid design system became essential to making AI-assisted work more predictable. My takeaway is to define constraints and acceptance criteria first, then work in small, reviewable steps: check AI-assisted findings against the original research, reuse established components, and verify layouts and states across screens. AI can accelerate exploration and execution; I remain responsible for the evidence, design decisions, and final quality.

Explore more projects