Dual-Agent Product Bot: Conversational AI with Memory Layer + Live Google Sheets Product Data Lookup

Problem:

A business needed a product assistant that could hold context-aware conversations across multiple turns while simultaneously querying live product data from Google Sheets, a single-agent setup couldn't cleanly separate conversation management from data retrieval without compromising response accuracy or memory continuity.

Solution:

Built a chat-triggered n8n two-agent pipeline, the first AI Agent (OpenAI + Simple Memory + Structured Output Parser) manages conversation context and processes the user's query with full session memory and structured output formatting; the parsed output is passed to the second AI Agent (OpenAI + Google Sheets read tool) which performs the actual live product data lookup against the connected sheet and returns accurate, context-aware product responses.

Result:

Delivered a dual-layer product assistant where memory management and data retrieval are handled by specialized agents, ensuring every response is both contextually aware of the conversation history and grounded in live product data from Google Sheets, producing more accurate and coherent answers than any single-agent approach.