Deploy Superset + MCP Server via Terraform, Integrate with Claude/GPT Chatbot

We're running a small, fixed-budget Proof of Concept to validate an AI-to-BI integration before committing to a larger engagement. Goal: Provide a working Terraform Infrastructure-as-Code setup that: Deploys an Apache Superset instance (v5.0+, required for native MCP support). Deploys and configures Superset's built-in MCP (Model Context Protocol) server alongside it. Demonstrates a working end-to-end connection between an MCP-compatible chatbot (e.g. Claude Desktop/Claude Code) and the running Superset instance - e.g. listing datasets, running a SQL query, or generating a chart via natural-language prompts. Reference: Please base your approach on Superset's official "Using AI with Superset" documentation (MCP setup, connection config, available tools): https://superset.apache.org/user-docs/using-superset/using-ai-with-superset/ Deliverables: - Terraform code (repo/zip) that provisions Superset + MCP server (cloud provider of your choice, reasonably documented). - Short setup/README instructions (how to apply Terraform, how to connect Claude or another MCP client). - A brief demo (screen recording or live call) showing the chatbot successfully querying/interacting with Superset via MCP.