SmartFolio

This innovative project aims to develop an intelligent, AI-powered agent that manages cryptocurrency portfolios on behalf of users. By leveraging advanced artificial intelligence and blockchain technology, the system provides a user-friendly, automated approach to crypto asset management. Key Features: Smart Wallet Integration: Each user is assigned a secure smart contract wallet, allowing for safe asset deposits and AI-controlled transactions within predefined limits. AI-Driven Investment Strategies: The core of the project is an AI agent that analyzes market data, technical indicators, and macroeconomic factors to provide tailored investment recommendations. Interactive User Interface: Users engage with the AI through a hybrid chatbot interface, combining conversational interactions with easy-to-use action buttons for executing trades. On-Chain Execution: The AI agent can perform cryptocurrency swaps and other DeFi operations directly on the blockchain, ensuring transparency and security. Gas Optimization: Initially, users will cover gas fees, with plans to implement gas sponsorship solutions in future iterations. Technology Stack: The project utilizes a cutting-edge tech stack, including LangChain for AI capabilities, blockchain-based smart wallets, and integration with various data sources for comprehensive market analysis.

User Experience: The system is designed to offer a seamless experience, from initial wallet connection and fund deposit to AI-suggested portfolio adjustments and on-chain trade execution. Users maintain control by reviewing and approving AI recommendations before any actions are taken. Future Potential: With plans for implementing account abstraction, advanced DeFi strategies, and social platform integrations, this project has significant potential for growth and enhanced functionality in the rapidly evolving cryptocurrency landscape.

1. Environment Setup: environment variables (OPENAI_API_KEY, CDP_API_KEY_NAME, CDP_API_KEY_PRIVATE_KEY etc) and configured the default networks

2. Agent Initialization: Initialized the agent with AgentKit by integrating action providers like WethActionProvider, erc20ActionProvider, and cdpWalletActionProvider to enable wallet management, token transfers, and other blockchain tasks.

3. LLM Integration: Used OpenAI's GPT-4o-mini model (ChatOpenAI) as the language model for decision-making, integrated with LangChain tools.

4. LangChain Tools: Leveraged LangChain's getLangChainTools to connect the agent with blockchain tools for executing on-chain operations.

5. Memory Management: Implemented a MemorySaver to persist risk and recommendation history, ensuring continuity in interactions

6. Custom Instructions: Provided detailed instructions (agentBaseInstruction, tradingAgentInstructions, etc.) to guide the AI agent's behavior for trading, risk recommendations, and portfolio management.

7. Market Data Integration: The agent fetched current market data through action providers like pythActionProvider for real-time price feeds.

8. Autonomous Actions: The agent autonomously analyzed portfolio risks, executed trades, and managed assets based on user input and market conditions.

9. Front-End Integration: The AI agent was integrated into a Nextjs-based front-end, allowing users to interact with it via a chatbot interface.