This is a Blockchain Event Prediction Market where users can utilize virtual currencies to predict various events and earn rewards.
UX designer
7.10 - 8.30
Users frequently spend considerable time collecting information before placing an order. And current AI agents lack specialization in event prediction. Additionally, some users face challenges in accessing AI tools.
Users find it difficult to efficiently collect relevant information, leading to delays in decision-making and order placement.
The lack of AI agents specialized in event prediction leaves users relying on general-purpose tools that do not meet their specific needs.
Many users also encounter difficulties accessing AI tools, such as ChatGPT, due to unfamiliarity with the platforms or complex interfaces.
How can we design an AI agent marketplace on the platform that offers high-quality, specialized models with clear, readable prediction processes, enabling users to explore more events and maximize their financial opportunities?
Users find it difficult to efficiently collect relevant information, leading to delays in decision-making and order placement.
The lack of AI agents specialized in event prediction leaves users relying on general-purpose tools that do not meet their specific needs.
Many users also encounter difficulties accessing AI tools, such as ChatGPT, due to unfamiliarity with the platforms or complex interfaces.
Provide professional information in a simple, jargon-free way, helping users easily understand and use the platform.
We achieved a 320% increase in the number of events users placed orders for, demonstrating the effectiveness of the platform's enhanced AI prediction features.
Additionally, users increased the total value of their orders by 180%, highlighting improved confidence in the pricing and event selection process.
AI Agent provides transparent prediction processes and diverse model options, enabling users to explore and select AI agents that maximize financial opportunities with ease and clarity.
This image shows the PredX platform, where users can use AI for predictions and leave reviews. The interface allows users to see details of the AI's prediction process, including content sources and reasoning. It highlights transparency, letting users understand how the AI reaches its conclusions, and encourages engagement through feedback.
It provides users with in-depth information about the AI model, including its performance metrics such as accuracy, reasoning level, explanation level, scalability, and credibility. The page also features a radar chart visualizing various evaluation factors, giving users a clear understanding of the AI agent’s strengths and areas for improvement.
Users can manage predictive events and placed orders. On the right, users can view and track prediction events, including details like cost, asset value, returns, and the AI agents used, all in one streamlined interface.
I conducted user interviews to understand their behaviors, challenges, and pain points during the event prediction process.
I analyzed competitor platforms to identify strengths and weaknesses in their AI prediction tools and user interfaces.
Based on insights from user feedback and competitor analysis, I iterated the design to improve usability, accessibility, and clarity.
Users prefer familiar events to minimize risk, but this limits event variety and leads to low platform liquidity. The lack of diversity results in lower income and higher risk. Making AI functionality easily accessible could help users quickly gather information, encouraging them to explore more events and improve liquidity.
Upon model upload, the system calculates a basic objective score.
Users then review the agent’s prediction process and submit their feedback.
A subjective score is generated from user reviews.
Provide professional information in a simple, jargon-free way, helping users easily understand and use the platform.
This is the complete process of my work, including interviews, competitor analysis, design solutions, and iterations.
