AI Lets Non-Coders Query Onchain Data
Crypto analytics platform Dune has launched its Master-Class Prompter (MCP), an AI tool that allows non-SQL users to query onchain data using natural language. The feature generates charts and statistics from simple text prompts, aiming to democratize access to complex blockchain data for a wider audience.
Dune's Master-Class Prompter (MCP) leverages the Model-Context Protocol, an open standard developed by Anthropic designed to create a universal interface between AI agents and external data sources. This client-server architecture abstracts the complexity of Dune's vast SQL databases, allowing different AI models to programmatically query on-chain data without needing custom, brittle integrations. This approach represents a shift from static, pre-built dashboards to dynamic query generation. The AI interprets a user's intent from natural language, identifies the relevant metrics and data tables, determines the necessary join paths, and then constructs the executable SQL query at runtime. This effectively introduces a live reasoning layer between the business user and the data platform. The push to democratize on-chain data addresses a significant gap; an MIT study estimated over 90% of raw blockchain data remains unutilized. Projections from AI firm Anthropic suggest that by 2025, as much as 70% of all blockchain queries could be handled by conversational NLP interfaces rather than by manually written code. This technology follows Dune's earlier release of DuneAI in November 2023, which first introduced a natural language engine to its platform. The company competes in a growing field of AI-driven crypto analytics tools, with platforms like Nansen, Glassnode, and Lore also offering AI-enabled features to interpret on-chain data and market signals. The rise of such AI assistants is reshaping data workflows. By automating the translation of business questions into code, these tools shift the focus for data professionals from routine SQL authoring to higher-value tasks like validating AI-generated insights, ensuring data quality, and building more robust underlying data models.
