Create a Paul Graham Essay Q&A System with OpenAI and Milvus Vector Database

Go to Workflow
661 views
Built by Cheney Zhang Cheney Zhang
Created on June 10, 2026

Description

Create a Paul Graham Essay Q&A System with OpenAI and Milvus Vector Database
How It Works

This workflow creates a question-answering system based on Paul Graham essays. It has two main steps:

Data Collection & Processing:
Scrapes Paul Graham essays
Extracts text content
Loads them into a Milvus vector store

Chat Interaction:
Provides a question-answering interface using the stored vector embeddings
Utilizes OpenAI embeddings for semantic search

Set Up Steps

Set up a Milvus server following the official guide
Create a collection named "my_collection"
Run the workflow to scrape and load Paul Graham essays
Start chatting with the QA system

The workflow handles the entire process from fetching essays, extracting content, generating embeddings via OpenAI, storing vectors in Milvus, and providing retrieval for question answering.

Nodes Used (9)

Default Data Loader
@n8n/n8n-nodes-langchain.documentDefaultDataLoader
Embeddings OpenAI
@n8n/n8n-nodes-langchain.embeddingsOpenAi
HTML
n8n-nodes-base.html
HTTP Request
n8n-nodes-base.httpRequest
Milvus Vector Store
@n8n/n8n-nodes-langchain.vectorStoreMilvus
OpenAI Chat Model
@n8n/n8n-nodes-langchain.lmChatOpenAi
Question and Answer Chain
@n8n/n8n-nodes-langchain.chainRetrievalQa
Recursive Character Text Splitter
@n8n/n8n-nodes-langchain.textSplitterRecursiveCharacterTextSplitter
Vector Store Retriever
@n8n/n8n-nodes-langchain.retrieverVectorStore