XPRT.ai — Advanced Data Annotation Mobile Platform

AI / ML Gamification Platform

XPRT.ai is a mobile-first, advanced data annotation tool designed for labelers and project managers in the AI/ML industry. By gamifying the annotation process, the app ensures high-quality, accurate datasets essential for reliable machine learning models. Available on iOS and Android, it allows users to annotate data across domains like retail, healthcare, sports, and more.

Annotate data quickly and earn rewards with the XPRT.ai Data Annotation Platform.

Key Features:

• Simplified Data Annotation: Users can label text, images, audio, and video effortlessly.
• Gamified Experience: Engages users with rewards and leaderboards, boosting accuracy and motivation.
• High Accuracy: Combines user selection with proprietary algorithms to deliver error-free results.
• Guided Onboarding: Personalizes the app experience and eases the learning curve for new users.

Design Process

Approach:
I adopted the User-Centered Design (UCD) process to focus on user needs throughout all design stages, ensuring a seamless and engaging experience.

My Design Process — UCD

Research

• Identified the primary goal: “Create a gamified AI training model app for both users and team managers.”
• Target Audience: AI/ML professionals, labelers, and project managers.
• Analyzed UI, UX, user flows, and information architecture of similar tools.

Insights from Research

• Users seek a simple, intuitive interface for efficient annotation.
• Gamification drives motivation and increases task completion rates.
• Managers want features for task scheduling, progress tracking, and performance insights.

Guided Introduction

We designed an onboarding experience to introduce app functionality to users before they start tasks, ensuring clarity and reducing friction.

How might we deploying the world class-AI and deliver a high-quality training data model through our app?

- Provided a guided introduction to personalize the app

- Creating a gamified interface for user engagement

- Simply giving accurate labeling and data annotation application for users

While creating a user profile, we created an introduction to the app for data annotation users where they can glimpse the app look at once before they perform the labeling tasks.

users can annotate images, video, audio, text data on the mobile platform.

Personas

Persona — 1

Persona — 2

Usability Testing

Tools:

• UsabilityHub for remote testing.
• Zoom for direct interviews with stakeholders and internal teams.

Questions Asked:

1. Can tasks be scheduled weekly or monthly?
2. How often should leaderboard points be updated?
3. What features do users value in a rewards-based annotation app?
4. Would notifications for new tasks improve engagement?

Empathy Map

Pain Points Identified:

• Difficulty in scheduling recurring tasks.
• Lack of real-time leaderboard updates.
• Limited customization options for rewards.
• Users wanted better notifications for new tasks and progress.

Defining the Problem

I decided to fix all the pain points that were important to Users and XPRT.ai. I redefined the pain points below.

Problem Definition and Solutions

Pain Points and Resolutions:

1. Scheduling Tasks: Introduced flexible task scheduling options (daily, weekly, or monthly).
2. Leaderboard Updates: Implemented real-time updates to keep users engaged.
3. Reward Features: Added customizable rewards based on task complexity and user achievements.
4. Notifications: Enabled push notifications for new tasks and milestones to ensure users stay informed.

Key User Challenges and Solutions

1. Discoverability of Category-Specific Tasks

Challenge:
Users struggled to start tasks in categories requiring certification (e.g., legal, medical), leading to confusion and reduced engagement.

Choosing interest — UI Screen

Solution:

• Introduced clear messaging in the app to guide users through uploading required certifications directly from the profile screen.
• Added a status tracker to show approval progress for restricted categories, ensuring transparency.
• Outcome: Improved clarity and reduced frustration for users attempting to access restricted tasks.

Lack of Gamification and Engagement

Challenge:
Users found tasks boring without gamified elements like leaderboards or rewards.

UI screens for Leaderboard and Gamification Profile

Solution:

• Implemented a leaderboard where users can view their performance, points, and category-specific rankings.
• Introduced super coins and a redemption page where users can exchange points for e-gift cards.
• Outcome: Increased user motivation and retention by fostering a sense of competition and reward.

Redeemption Page

3. Irrelevant Questions in Tasks

Challenge:
Users occasionally received unrelated or random questions, causing frustration during annotation tasks.

4: First-time users can view demo introduction to the mobile application in the Xprt.ai app

Challenge:
New users lacked guidance on navigating and using the XPRT.ai app effectively, leading to confusion and potential disengagement.

Solution:

• Designed a demo introduction screen that launches upon the first login, showcasing key features and functionality through a simple, interactive walkthrough.
• Highlighted steps like category selection, certification upload, task navigation, and gamified elements to ensure clarity.

Prototyping and Validation

After addressing these challenges, I developed Hi-Fi prototypes using Adobe XD. The clickable prototypes underwent usability testing with five new participants. Feedback confirmed that the solutions resolved key issues and improved user engagement.

Below are the final Hi-Fi mockups illustrating the implemented design solutions.

Validating the Prototype

During the prototype validation phase, some usability issues were identified, and I focused on refining the following areas:

1. Reporting Questions During Tasks

Problem:
Users encountered difficulties with certain task questions but lacked a way to flag issues. This led to frustration and a negative user experience.

Solution:

• Added a “Report” button for each question on the task UI screens.
• This feature allows users to report conflicts or unclear questions in real-time without disrupting their workflow.
• Outcome: Improved user satisfaction and streamlined issue resolution.

2. Difficulty Levels and Badge System

Problem:
Users found it unclear how expertise levels and badges were assigned for different categories like medical, legal, or sports. The initial badge designs did not effectively convey progression or achievement.

Solution:

• Introduced category-specific difficulty levels with clear criteria for earning badges (e.g., “Expert” vs. “Regular” users).
• Collaborated daily with the team to refine badge designs through several iterations, ensuring clarity and alignment with user expectations.
• Final designs were approved by stakeholders and review teams after usability testing.

Outcomes

The Report Button resolves task conflicts seamlessly, enhancing user experience, while the Badge System motivates users by recognizing their progress and expertise.

By integrating user feedback and gamification, XPRT.ai ensures efficient data annotation with improved usability, fostering engagement and productivity.

This project showcases my ability to design engaging, user-focused solutions in the AI/ML industry.