This n8n workflow automates the transformation and synchronization of your organization’s technology radar data across Google Sheets, Google Docs, MySQL, and Pinecone vector databases. It enables both structured SQL querying and retrieval-augmented generation (RAG) document search, allowing internal AI agents to provide strategic technology adoption insights for multiple companies in real time. It includes an API endpoint for conversational queries, intelligently routing them to the appropriate AI agent (SQL or RAG) and applying guardrails before returning responses.
Features:
- Google Sheets → MySQL Sync: Periodically imports and updates tech radar records into a structured MySQL database for SQL-based AI agents.
- Google Sheets → Google Docs → Vector DB: Converts spreadsheet rows into paragraphs, updates a Google Doc, and ingests it into a Pinecone vector database for RAG-based AI agents.
- Intelligent AI Routing: Uses an LLM to decide whether a user query is best handled via SQL lookup or RAG search.
- Dual AI Agent Workflows: Separate subworkflows for SQL and RAG allow flexible, accurate responses based on the type of data needed.
- Webhook Chat Endpoint: Provides a simple API to query the radar via chat-like input, returning validated AI-generated answers.
- Built-In Guardrails: Ensures responses are aligned with strategic direction, restricted to relevant tech adoption guidance, and never leak sensitive prompts.
- Fully Extensible Setup: Supports multiple companies, strategic directions, and easy customization for different tech radar data sources or vector databases.