FlowKit
FRENSee the packs

4 n8n workflows · JSON + guide

RAG Assistant Pack

A chatbot that answers from your own documents, with citations

Four n8n workflows that turn your PDFs and Notion pages into a documentation assistant: ingestion, chunking, embeddings, storage in Supabase pgvector, then a chat that answers with source citations — plus an API to embed it anywhere.

€79incl. VAT · one-time

Buy the RAG Assistant Pack79 €Secure Stripe checkout · Instant ZIP delivery · 14-day money-back guarantee

Instant download · Setup guide in French — workflows are language-agnostic

The pack

What this pack changes

Your procedures, contracts, documentation and internal notes sleep in PDFs nobody re-reads. This pack builds a complete RAG (Retrieval-Augmented Generation) pipeline on your own infrastructure: documents are chunked and vectorized into Supabase, and an n8n chatbot answers your team's questions using those contents exclusively, citations included. No out-of-scope hallucinations: when the answer isn't in your documents, it says so.

Who it's for

  • SMBs with an internal knowledge base (procedures, HR, quality) to make searchable
  • Consultants who want to offer clients a documentation assistant
  • Support teams answering the same questions from the same PDFs

What's inside

  • 4 n8n workflows as JSON files, importable in two clicks
  • The complete pgvector setup SQL (table, index, match_documents function)
  • Step-by-step setup guide (in French — workflows are language-agnostic)
  • Citation-aware answer prompt, ready to adapt to your tone

Requirements

  • n8n ≥ 1.60 (cloud or self-hosted)
  • An OpenAI API key (embeddings + generation)
  • A free Supabase project with the pgvector extension enabled
  • Optional: a Notion workspace for the nightly sync

The contents, file by file

The 4 workflows, in detail

01 · RAG PDF ingestion

Ingest PDFs into Supabase pgvector with n8n

First link of a RAG pipeline: this workflow exposes a webhook that accepts your PDFs, extracts the text, splits it into chunks, computes embeddings and inserts everything into a Supabase pgvector table. One curl command indexes a document.

See how it works node by node
Pipelineingestion-pdf-supabase-pgvector.json · 6 nodes
  1. Dépôt de PDF (Webhook)triggerWebhook
  2. Insérer dans la base vectorielleSupabase vector store
  3. Confirmer l'ingestionWebhook response

Attached AI sub-nodes

  • Chargeur de PDFDocument loader
  • Découpage du texteText splitter
  • Embeddings OpenAIOpenAI embeddings

02 · RAG chatbot

Build a RAG chatbot with citations in n8n

The heart of the RAG Assistant Pack: a public n8n chat that queries your Supabase vector store and answers with source citations. Six nodes — proof that production RAG requires neither a framework nor code.

See how it works node by node
Pipelinechatbot-rag-citations.json · 6 nodes
  1. ChattriggerChat (trigger)
  2. Assistant documentaireRAG chain (Q&A)
  3. Base vectorielle SupabaseSupabase vector store

Attached AI sub-nodes

  • Modèle OpenAIOpenAI model
  • RetrieverRetriever
  • Embeddings OpenAIOpenAI embeddings

03 · Notion RAG sync

Sync Notion to a vector store with n8n

Your documentation lives in Notion? This workflow indexes it automatically: every night at 2 am, it walks a Notion database, chunks and vectorizes the content, then refreshes the vector store — your RAG chatbot stays in sync with today's truth.

See how it works node by node
Pipelinesync-notion-base-vectorielle.json · 9 nodes
  1. Toutes les nuits à 2htriggerSchedule (cron)
  2. Pages de la base NotionNotion
  3. Préparer le contenuSet (fields)
  4. BoucleLoop (batches)
  5. Insérer dans la base vectorielleSupabase vector store
  6. Notifier la fin sur SlackSlack

Attached AI sub-nodes

  • Chargeur de texteDocument loader
  • Découpage du texteText splitter
  • Embeddings OpenAIOpenAI embeddings

04 · RAG API

Expose a question-answering API over your documents with n8n

The n8n chat can't be embedded everywhere — an API can. This workflow exposes a POST /ask endpoint: send {"question": "…"} and receive the answer grounded in your documents as structured JSON, ready to display in your site, CRM or internal tool.

See how it works node by node
Pipelineapi-question-reponse-rag.json · 10 nodes
  1. POST /ask (Webhook)triggerWebhook
  2. Question valide ?IF (condition)
  3. Répondre depuis les documentsRAG chain (Q&A)
  4. Base vectorielle SupabaseSupabase vector store
  5. Formater la réponseSet (fields)
  6. Réponse JSONWebhook response
  7. Erreur 400Webhook response

Attached AI sub-nodes

  • Modèle OpenAIOpenAI model
  • RetrieverRetriever
  • Embeddings OpenAIOpenAI embeddings

RAG Assistant Pack — €79 incl. VAT

One-time payment, instant download, updates included. 14-day money-back guarantee: if a workflow doesn't import, we refund you.

Or get the Complete Bundle: all 3 packs for €179 instead of €227.

Buy the RAG Assistant Pack79 €Secure Stripe checkout · Instant ZIP delivery · 14-day money-back guarantee

FAQ

Frequently asked questions

Which document formats are supported?

The ingestion workflow handles PDFs natively (text extraction included). The Notion workflow covers your pages and databases. For other formats (DOCX, HTML), the guide shows how to add an extraction node upstream — the vectorization chain stays identical.

Why Supabase instead of Pinecone or Qdrant?

Because Supabase's free tier easily handles thousands of pages, pgvector is open source, and your data stays in a regular Postgres you can query, back up and migrate. The workflows use n8n's official Supabase nodes.

Can the chatbot make up answers?

The included prompt constrains the model to answer only from retrieved passages and to cite the source of every claim. If nothing relevant is found, it explicitly says it doesn't know. You can tighten or relax that instruction.

How many documents can I index?

Supabase's free tier (500 MB) typically absorbs several thousand pages split into 1,000-character chunks. Beyond that, Supabase's paid plan (~$25/month) pushes the limit much further — without changing the workflows.