[ChainOpera AI]
Fortune Telling AI Agent
Designing an emotionally engaging AI experience for Web3 communities
ChainOpera AI -Soulchain
Fortune Telling Agent for web 3
[Overview]
ChainOpera AI (CO) sought to create a new AI Agent that brings entertainment and personality into its Web3 community, turning daily AI interactions into delightful rituals. As the sole UIUX designer, I transformed the product requirements into a cohesive conversational flow and a visual identity system inspired by tarot mysticism and CO’s brand aesthetics to engage the Web3 community and express the brand’s mystical yet futuristic tone.
[Highlights]
Enhanced community engagement through entertainment UX
Developed a fortune-telling experience that encouraged daily interactions, strengthening ChainOpera’s brand presence within its decentralized ecosystem.
Drove AI-powered design production
Created a complete set of 78 tarot cards blending the traditional Rider–Waite symbolism with ChainOpera AI’s brand identity.
Reached 130k+ users in the first month
The agent quickly became one of the most interacted-with entertainment experiences in the ChainOpera ecosystem, strengthening community engagement through playful AI design.
Problem & Opportunity
Most AI agents focus on utility. The challenge was to design one that feels personal, magical, and emotionally rewarding, while staying aligned with ChainOpera’s brand voice and Web3 aesthetic.
Users in the Web3 space engage with projects that feel personalized, playful, and visually symbolic.
There was a gap between utility-driven agents and emotionally engaging experiences
Chat Flow Mapping
User Flow
Since this agent is implemented within a chat-based interaction, the user flow differs from traditional interface navigation. Instead of defining screens and buttons, it serves as a guideline for how the AI agent should lead the conversation—from greeting the user, identifying their intent, and guiding card selection, to delivering explanations and suggesting follow-up actions.
Unlike traditional tarot readings that involve complex spreads, this AI agent introduces a simplified tarot experience, allowing users to choose 1, 3, or 5 cards directly within the chat. This streamlined approach lowers the learning curve, making tarot reading more accessible while preserving its symbolic depth and ritualistic feel.
Since ChainOpera’s core agents focus on utility use cases in the Web3 space, this entertainment agent also encourages users to explore other ChainOpera agents during the conversation, fostering cross-agent engagement within the ecosystem.
UI Screens
Greeting
Interactive Card Draw System
Advanced Tarot
AI-Powered Visual System
Reimagined Rider–Waite symbolism in ChainOpera’s visual language (created with Dreamina)
When designing the tarot cards with AI, the main challenge was to preserve the original symbolic meaning of the Rider–Waite Tarot while ensuring stylistic consistency across all 78 cards. So the first step was to deeply understand the meaning behind each card, then select four representative ones to experiment with different visual styles:
AI-generated imagery often comes with several limitations:
Inaccurate details — Facial features or hands can appear distorted (e.g., too many fingers or blurred expressions)
Difficulty in precise visual control — Describe each card’s complex imagery accurately through prompts alone
Inconsistent styles — Maintaining a unified visual tone across generations is difficult
Niche symbolic elements — Tarot suits such as Pentacles, Swords, Cups, and Wands feature small, intricate motifs that AI models may fail to capture correctly
To overcome these limitations, I experimented with multiple AI tools and iterative workflows:
Used Dreamina for initial concept generation, exploring composition, lighting, and symbolic direction.
Switched to LoveArt AI for refining details and correcting distortions, especially facial expressions and object accuracy.
Manually curated and adjusted outputs to ensure consistency in color palette and symbolism across the full set.
