Updated
Updated · AWS Blog · Jun 1
Amazon Quick Connects KDB-X via MCP, Turning 10 Market Queries Into Natural-Language Analysis
Updated
Updated · AWS Blog · Jun 1

Amazon Quick Connects KDB-X via MCP, Turning 10 Market Queries Into Natural-Language Analysis

1 articles · Updated · AWS Blog · Jun 1
  • Amazon Quick now links to KDB-X time-series databases through Model Context Protocol, letting financial analysts ask conversational questions and get market insights without writing SQL or q queries.
  • Amazon Bedrock AgentCore Gateway sits between Quick and the MCP server as the authentication and routing layer, while Amazon Cognito supplies OAuth credentials required for the Quick connector.
  • On the backend, KDB-X and its MCP server run on an EC2 instance—t2.medium or larger—with KDB-X listening on port 5000 and tools such as run_sql_query, hybrid_search and similarity_search exposed to Quick.
  • The walkthrough shows the setup can answer tasks such as sampling 10 stocks, calculating total trading volume, finding intraday peaks and generating price charts from time-stamped trade data.
  • Amazon positions the pattern as broader than finance, saying the same MCP-based approach can simplify access to time-series insights for IoT monitoring, DevOps dashboards and other data-heavy workloads.
As Amazon's AI agent enters the 'Cowork' market, can its open-standard approach succeed where its previous productivity tools failed?
By replacing expert query languages with AI, are financial firms trading precision and security for simple convenience?
With AI now handling complex data queries, what is the new role for the human analyst in a world of agentic intelligence?