RAG Chatbot with Supabase + TogetherAI + Openrouter
VerifiedIngests documents into Supabase and powers a RAG chatbot via OpenRouter.
What this workflow does
This automation ingests Google Docs content into a Supabase vector database using Together AI embeddings and enables retrieval-augmented chat responses through OpenRouter models.
It is designed for teams needing a private knowledge chatbot that answers questions based solely on uploaded documents without external data leakage.
Who is this for?
Developers and small teams building internal knowledge assistants or document-based Q&A tools. Ideal for technical users comfortable with API keys and vector databases.
What problem it solves
Manually searching long documents or knowledge bases is slow and error-prone. This workflow automates retrieval-augmented generation so users can chat naturally with their own content.
Live workflow preview
Interactive canvas of every node and connection — scroll and click to explore. Powered by n8n's preview.
Open the template on n8n to import and run it. View source template →
What it automates
Company policy Q&A
Upload HR or compliance docs once; team members ask questions via chat and receive grounded answers from the stored embeddings.
Product documentation bot
Convert product manuals in Google Docs into a searchable chatbot for support staff or customers.
Research note assistant
Embed research notes or reports so analysts can query specific findings without rereading entire files.
How the workflow works
The 6 nodes in this automation, in order.
- 1HTTP RequesthttpRequest
- 2Google DocsgoogleDocs
- 3Supabasesupabase
- 4Codecode
- 5Basic LLM Chain@n8n/n8n-nodes-langchain.chainLlm
- 6OpenRouter Chat Model@n8n/n8n-nodes-langchain.lmChatOpenRouter
Apps & integrations used
How to set up RAG Chatbot with Supabase + TogetherAI + Openrouter
- 1Import both workflow JSON files into n8n
- 2Add Google service account credentials to the Google Docs node
- 3Create the Supabase embed table and add connection credentials
- 4Insert TogetherAI API key into the HTTP Request embedding nodes
- 5Configure OpenRouter credentials in the Basic LLM Chain node
- 6Run the first workflow once to ingest documents, then activate the chat workflow
How to customize this workflow
- →Swap TogetherAI embeddings for another provider by editing the HTTP Request URL and payload
- →Change the chat trigger from chatTrigger to telegramTrigger or another messaging node
- →Modify chunk size or splitting logic inside the code node
- →Add extra Google Docs nodes before the splitter to ingest multiple sources
RAG Chatbot with Supabase + TogetherAI + Openrouter: pros & cons
Pros
- +One-time ingestion workflow keeps embeddings fresh
- +Supabase vector search is fast and built-in
- +OpenRouter gives easy model switching
- +Clear separation between setup and chat flows
Cons
- –First workflow must be run manually each time content changes
- –Chunking relies on custom code rather than a dedicated splitter node
- –Requires multiple paid API services
Frequently asked questions
It ingests Google Docs into Supabase embeddings and lets you chat with that content using a RAG pipeline.
User reviews
Verified reviews from the community shape this listing's rating.
Loading reviews…
Similar workflows
Other ai & llm automations worth a look.