AI Agents

Generate citations for OpenAI responses using RAG file retrieval

19 nodes 308 172 Automatic trigger
Download

Workflow Description

This automation receives user queries through a chat interface, retrieves relevant documents from stored memory, sends them to OpenAI along with the question, and returns answers with accurate citations pointing to original sources.

How it works

  1. 1.Receive question through chat interface
  2. 2.Retrieve related files from stored memory
  3. 3.Send context and question to OpenAI
  4. 4.Process response and attach citations
  5. 5.Return result to user

Use cases

  • Answer questions about company documents while citing exact sources
  • Build an AI assistant that traces every fact to its original reference

Requirements

  • Valid OpenAI API key
  • Collection of documents or files stored in memory system

Service Value

Ideal as a smart automation service combining integrations and AI to produce ready-to-use results.

Apps Used

Memory Chat Trigger OpenAI

Details

Trigger Automatic trigger
Nodes 19
Apps 3
Views 308
Downloads 172

How to Use

  1. 1.Click "Download Template"
  2. 2.Open your n8n dashboard
  3. 3.Go to Workflows > Import from File
  4. 4.Select downloaded file and configure credentials

Nodes Used (19)

/

Aggregate

Aggregate

#1

Window Buffer Memory

Memory Buffer Window

#2

Sticky Note4

Sticky Note

#3

Sticky Note

Sticky Note

#4

Create a simple Trigger to have the Chat button within N8N

Chat Trigger

#5

OpenAI Assistant with Vector Store

OpenAI

#6

Sticky Note1

Sticky Note

#7

Get ALL Thread Content

HTTP Request

#8

Sticky Note2

Sticky Note

#9

Split all message iterations from a thread

Split Out

#10

Split all content from a single message

Split Out

#11

Split all citations from a single message

Split Out

#12

Retrieve file name from a file ID

HTTP Request

#13

Regularize output

Set

#14

Sticky Note3

Sticky Note

#15

Sticky Note5

Sticky Note

#16

Sticky Note6

Sticky Note

#17

Optional Markdown to HTML

Markdown

#18

Finnaly format the output

Code

#19