n8n Document Q&A Webhook with OpenAI Chat + Vector Search

Answer support questions from your PDFs with an n8n Document Q&A Webhook (OpenAI Chat + Vector Search)

Turn your existing documents into a searchable knowledge base and expose a secure, header-authenticated webhook that answers customer support questions using OpenAI chat—with optional retrieval from your uploaded PDF/CSV/TXT content.

What this workflow does

• Ingest documents: Receives uploaded PDF, CSV, or TXT files via an n8n form and replaces the current in-memory knowledge base with the new content.
• Build a vector index: Splits files into chunks, generates embeddings with OpenAI, and stores vectors in an in-memory vector store.
• Secure Q&A webhook: Handles a POST request on a header-authenticated webhook containing a user message and a session_id.
• Chat-first support agent: Uses an OpenAI-based support agent with a system prompt and 15-message conversation memory to answer directly when possible.
• Retrieve facts when needed: If the agent requires document grounding, it runs a retrieval agent that searches the in-memory vector store and returns sourced information only.
• Returns plain text: Formats the final response as plain text and sends it back in the webhook response.

Use cases

• Enable SaaS support teams to answer billing, policy, or product questions from internal documentation.
• Automate knowledge-based responses for onboarding FAQs uploaded as PDFs or text files.
• Provide consistent answers by using a session_id for short, session-based conversation context.

Technical details

• Workflow category: n8n automation workflow
• Key nodes/steps: set, webhook, sticky note, form trigger, n8nn8n-nodes-langchainagent (OpenAI support agent + retrieval), and respond to webhook.
• Integrations/requirements: OpenAI for embeddings and chat model; an HTTP Header Auth credential for the webhook.

SEO tip: Ideal for teams building an “OpenAI document Q&A” system inside n8n with “vector search” over “PDF/CSV/TXT”.