By leveraging LLM and analytics mechanisms to provide natural language queries, MoonDash will simplify on-chain data analysis through real-time insights and intuitive, customizable dashboards.
We are solving the challenge of accessing and analyzing data within different blockchain ecosystems by building LLM-powered analytics platforms. MoonDash will allow users to interact with on-chain data using natural language queries rather than requiring technical expertise or learning complex ontologies. Users will be able to easily compose custom dashboards, empowering both technical and non-technical participants to gain real-time insights from blockchain data.
The integration of an LLM provides users with the flexibility to ask questions about networks in simple terms. For example, a user could ask:
- “What were the largest ADA transactions in the last 24 hours?”
- “How many accounts hold over 200,000 XML?”
- “Plot a bar chart showing monthly smart contract deployments in Solana in the last year.”
Ask questions about blockchain data in plain English, powered by advanced LLM technology
Create personalized visualizations and monitoring tools for your specific needs
Compare and analyze data across different blockchain ecosystems
Comprehensive tracking of our blockchain analytics initiatives across multiple ecosystems
Status: Fully Funded
Total Funding: USD150,000
Timeline: Feb 2025 - Dec 2025
Code: GitHub
Status: Funding Needed
Status: Funding Needed
Status: Funding Needed
Status: Funding Needed
Status: Funding Needed
Status: Funding Needed
Advanced capabilities powered by artificial intelligence
Large Language Models (LLMs) offer powerful natural language understanding capabilities but face challenges with factual consistency and domain-specific knowledge. Their main advantages include flexible natural language processing, contextual understanding, and the ability to generate human-like responses. However, they can often produce hallucinations or inconsistent outputs when dealing with specialized domain knowledge.
Knowledge Graphs (KGs) excel at representing structured domain knowledge and supporting precise querying but lack natural language understanding. Their strengths include explicit relationship modeling, formal reasoning capabilities, and guaranteed factual consistency within their domain. However, they require specialized query languages and lack the flexibility to handle natural language inputs and results.
A Knowledge Graph Enhanced LLM combines these approaches to mitigate their respective limitations. By integrating LLM responses in formal ontologies and knowledge graphs, it is possible to ensure factual accuracy while maintaining natural language interaction capabilities. In the context of blockchain analytics, this means MoonDash can leverage the LLM's ability to interpret user queries while ensuring responses are based on accurate blockchain data and relationships defined in our ontology and KG.