Uniicon

• 

  Lander page

• 

  Icon generation page

Inspiration

Ever noticed that 3d animated emojis are getting popular in modern website. There are no fixed platforms to build this using an agentic approach to both building and getting paid at the same time.

What it does

Uniicon is an AI-powered platform that generates stunning 3D animated icons from simple text descriptions. The system employs a sophisticated multi-agent pipeline where:

1. Extract Agent (Amazon Bedrock) analyzes user input to extract key visual elements
2. Icon Generator (OpenAI GPT) creates high-quality 3D icons based on extracted data
3. Background Cleaner (Replicate API) removes backgrounds for clean results
4. Interpret Agent (Amazon Bedrock Nova Lite) analyzes the generated icon for animation planning
5. Planner Agent (Amazon Bedrock) creates detailed animation strategies
6. Animation Engine (RunwayML) transforms static icons into engaging video animations

The platform automatically mints generated icons as NFTs on Base Sepolia and supports crypto payments through X402 protocol integration. Users can also integrate with the system through LangChain agents or Claude's Model Context Protocol (MCP) server.

How we built it

The project was built using a modern tech stack:

• Frontend: Next.js 15 with React 19 and TailwindCSS for a responsive, modern UI
• AI/ML Pipeline: Amazon Bedrock for multi-agent orchestration, OpenAI for icon generation, RunwayML for animation
• Blockchain: Base Sepolia testnet with Wagmi, Viem, and RainbowKit for wallet connectivity
• Payment System: X402 protocol for seamless USDC transactions with automatic revenue distribution
• Storage: Pinata IPFS for decentralized storage and metadata management
• Agent Integrations: LangChain for autonomous agent interactions and MCP server for Claude integration

The core architecture follows a microservices approach where each AI agent has a specific responsibility in the pipeline. The system uses streaming responses for real-time processing and implements proper error handling throughout the workflow.

Challenges we ran into

1. Multi-Agent Coordination: Designing a pipeline where different AI models work together seamlessly was complex. Each agent needed to understand the output format of the previous agent and provide appropriate input for the next one.

2. Animation Quality Control: Ensuring consistent quality in AI-generated animations required extensive prompt engineering and fine-tuning of the RunwayML integration.

3. Blockchain Integration: Implementing automatic NFT minting while maintaining user experience was challenging, especially handling gas fees and transaction failures gracefully.

4. Payment Flow: Integrating X402 protocol for crypto payments required careful handling of wallet connections, payment verification, and revenue distribution.

5. Real-time Processing: Managing long-running processes (2+ minutes) while keeping users engaged required implementing proper progress indicators and background processing.

6. Agent Framework Integration: Building both LangChain and MCP server integrations required understanding different agent architectures and payment handling mechanisms.

Accomplishments that we're proud of

1. Successful Multi-Agent Pipeline: Created a sophisticated AI system where multiple agents work together seamlessly to produce high-quality animated icons.

2. Blockchain Integration: Successfully implemented automatic NFT minting with proper metadata storage on IPFS, making every generated icon a tradeable digital asset.

3. Payment Innovation: Integrated X402 protocol for seamless crypto payments, creating a true pay-per-use model with automatic revenue distribution.

4. Agent Ecosystem: Built comprehensive agent integrations supporting both LangChain and Claude MCP, making the platform accessible to various AI agent frameworks.

5. User Experience: Despite the complex backend, maintained a simple, intuitive interface that hides technical complexity from users.

6. Demo Video: Created a comprehensive demo showcasing all features and integrations, making the project accessible to both technical and non-technical audiences.

What we learned

1. AI Agent Architecture: Learned how to design and implement multi-agent systems where each agent has specialized capabilities and can work together effectively.

2. Blockchain Development: Gained deep understanding of smart contract development, NFT minting, and decentralized storage solutions.

3. Payment Protocol Integration: Learned to implement X402 protocol for crypto payments, including wallet client management and payment verification.

4. Agent Framework Design: Understood how to build tools that can be integrated with different AI agent frameworks while maintaining payment handling.

5. AWS Bedrock: Gained expertise in using Amazon Bedrock for multi-agent orchestration and streaming responses.

6. Animation AI: Learned to work with RunwayML for AI-generated animations and prompt engineering for consistent results.

What's next for Uniicon

1. Mainnet Deployment: Migrate from Base Sepolia testnet to mainnet for production use with real value transactions.

2. Advanced Customization: Add more customization options for animation styles, durations, and visual effects.

3. Marketplace Integration: Build a marketplace where users can buy, sell, and trade their generated animated icons as NFTs.

4. Mobile Application: Develop native mobile apps for iOS and Android to make icon generation accessible on mobile devices.

5. Enterprise Features: Add bulk generation, API access, and custom branding options for enterprise customers.

6. Community Features: Implement user galleries, sharing capabilities, and community voting on best icons.

7. Advanced Agent Integrations: Expand agent framework support to include more platforms and add more sophisticated agent capabilities.

8. Performance Optimization: Reduce generation time through model optimization and parallel processing improvements.

Built With

• hardhat
• langchain
• nextjs
• react
• solidity