What it does
An n8n workflow that connects several LLMs to one AI chain and uses them in turn. It drafts a reply to a customer complaint with one model, checks the reply with a sentiment analysis step, and tries the next model in the list if the reply is not satisfactory. The loop ends when a reply passes or every model has been used.
Use cases
- 01Fall back to another model when a reply fails the check
- 02Draft replies to customer complaints for review
- 03Compare how different LLMs answer the same message
Connects
Basic LLM Chain · LangChain Code · OpenAI Chat Model · Sentiment Analysis