BlockIQ

Inspiration

The blockchain ecosystem is growing rapidly, but most data is locked behind complex APIs or raw explorers that are hard to interpret. We wanted to democratize blockchain insights, making wallet activity and transaction data understandable and actionable for everyone, not just developers. Inspired by the potential of AI agents and the Model Context Protocol (MCP), we aimed to create a conversational platform where AI can analyze blockchain data and generate interactive visualizations automatically.

What it does

BlockIQ is an AI-powered blockchain analytics platform that allows users to:

• Query wallet balances, transaction histories, and activity patterns using natural language.
• Run comprehensive analyses with our executeSpiderAnalysis tool, including top recipients, transaction trends, and first/last transactions.
• Generate interactive visualizations in real time, including bar charts, line charts, pie charts, and more using our Graph Protocol.
• Leverage multi-agent orchestration to handle complex queries with provider-agnostic AI models (OpenAI, Google, Anthropic, Groq, Hugging Face).

In short, BlockIQ turns complex blockchain data into intuitive insights that anyone can understand.

How we built it

• Framework & Architecture: Built on the ADK-TS framework, orchestrating multiple AI agents and integrating a custom MCP server.
• Backend: Node.js + Next.js API routes manage data fetching, caching, and AI interactions.
• Caching: Implemented smart incremental caching to minimize API calls and deliver near-instant responses for large wallets.
• AI Models: Supports multiple providers and models, configurable via the settings panel.
• Frontend: Next.js + React + TailwindCSS + Framer Motion for a responsive, animated, and interactive user experience.
• Visualization: AI automatically generates graph JSON via Graph Protocol, which our frontend renders into dynamic charts with tooltips and animations.

Challenges we ran into

• Handling massive transaction histories efficiently without overwhelming the API.
• Designing self-correcting AI workflows to automatically debug and rerun analysis scripts.
• Ensuring real-time interactivity for users while maintaining backend reliability and caching correctness.
• Balancing technical complexity with accessibility for non-technical users.

Accomplishments that we're proud of

• Successfully orchestrated multi-agent AI analysis over thousands of transactions in under 2 seconds.
• Built a self-healing MCP tool that automatically detects and fixes code errors during analysis.
• Designed Graph Protocol to let AI autonomously choose chart types, colors, titles, and insights from raw blockchain data.
• Created a user-friendly interface that makes blockchain analytics approachable for everyone.

What we learned

• Multi-agent AI systems can meaningfully interact with complex datasets if structured properly.
• Incremental caching and smart preloading are essential for high-performance AI applications with large data volumes.
• Visual storytelling with AI insights greatly improves usability and engagement.
• Balancing technical depth with clarity for beginners is key for hackathon projects.

What's next for BlockIQ

• Mainnet support: Extend beyond the Stacks testnet for real-world usage.
• Cross-chain analytics: Integrate Ethereum, Solana, and other chains.
• Advanced AI collaboration: Enable agents to collaborate and provide predictive insights on user portfolios.
• Expanded visualization types: Add more interactive and dynamic charting options based on AI-detected patterns.

Built With

• adk-ts-framework
• framermotion
• next.js
• node.js
• openai/anthropic/grok/huggingface
• react
• self-graph-protocol
• smart-caching
• stack-api
• tailwindcss
• typescript
• vscode