Claude Haiku Webhook Chatbot Workflow (n8n) for Service FAQs

Answer Service FAQs with an Anthropic Claude Haiku Webhook Chatbot (n8n)

Install this Claude Haiku Webhook Chatbot workflow for n8n to power a simple POST /chat endpoint that responds to customer questions—focused strictly on Ali İlhan’s services and tech stack. It returns a clean JSON reply and keeps short-term context per user session.

What this workflow does

• Receives chat requests via a webhook at /chat using a POST body containing message and optional sessionId.
• Validates input and rejects empty messages to prevent pointless model calls.
• Uses an n8n LangChain agent backed by Anthropic Claude Haiku to generate answers.
• Enforces a strict system prompt so the agent only answers service / project / tech stack questions and refuses everything else.
• Maintains session-based memory with buffer window memory so follow-up questions work naturally per sessionId.
• Returns a JSON response containing reply, success, and timestamp.
• Responds with CORS + JSON headers for smooth frontend integration.

Use cases

• Quick FAQ chatbot on a website section dedicated to Ali İlhan’s services and technology stack.
• Automation engineers need a lightweight webhook-driven AI Q&A endpoint for SaaS or internal tools.
• Route customer questions to accurate, constrained answers without opening up the model to unrelated topics.

Technical details

• Webhook node to expose a /chat endpoint
• LangChain Agent (n8nn8n-nodes-langchainagent)
• Anthropic Claude Haiku chat model (n8nn8n-nodes-langchainlm chat anthropic)
• Input handling with code and request validation
• Webhook response using respond to webhook and JSON formatting

Setup: Add an Anthropic API credential, select it in the Claude Haiku node, deploy the workflow, and call the production webhook URL with Content-Type: application/json including message (and optional sessionId).