Amazon Quick + KDB-X MCP lets analysts query time-series data via plain English
No more complex SQL: ask questions like 'Show me stock volatility in the last hour'
Amazon Quick, Amazon's generative AI-powered BI service, now integrates with KDB-X time-series databases via the Model Context Protocol (MCP). This eliminates the need for complex database queries, letting financial analysts ask conversational questions like 'Compute the volatility of AAPL over the last 5 minutes' and receive actionable insights from high-frequency market data. The architecture uses a KDB-X MCP server deployed on an Amazon EC2 instance (t2.medium or larger). Amazon Quick translates natural language into SQL, which is passed to the MCP server, then executed against the KDB-X database.
To secure and route these requests, the solution relies on Amazon Bedrock AgentCore Gateway—a single access point that authenticates users via Amazon Cognito and validates inbound requests before forwarding them to the MCP server. The MCP server exposes domain-specific tools such as hybrid_search, run_sql_query, and similarity_search, enabling tasks like semantic search on SEC filings or real-time market data analysis. While demonstrated for financial market intelligence, the same pattern can be applied to IoT sensor monitoring and DevOps dashboards, democratizing access to time-series insights across any industry.
- Amazon Quick MCP integration lets non-technical users query KDB-X time-series databases in plain English, removing the need for SQL expertise.
- The setup uses an EC2-hosted KDB-X MCP server, Amazon Bedrock AgentCore Gateway for routing, and Amazon Cognito for identity management.
- Tools include run_sql_query, hybrid_search, and similarity_search—enabling volatility calculations, market data queries, and semantic SEC filing search.
Why It Matters
Democratizes high-frequency time-series analysis, letting analysts ask data questions instantly without writing queries.