AI Agents

Create AI-ready vector datasets for LLM students using Gemini and Pinecone

21 nodes 221 126 Automatic trigger
Download

Workflow Description

A workflow that prepares AI-ready vector datasets for students by splitting documents, generating embeddings with Gemini, and storing them in Pinecone for semantic search, with optional X sharing for community engagement.

How it works

  1. 1.Trigger manually and load documents using the Document Data Loader
  2. 2.Split text into overlapping chunks with the Recursive Character Text Splitter
  3. 3.Generate embeddings via Google Gemini and store vectors in Pinecone
  4. 4.Extract structured information using LLM Chain with output parsing
  5. 5.Publish results to X for community sharing and feedback

Use cases

  • Building curated training datasets for machine learning and LLM courses
  • Creating searchable knowledge bases for student research and learning projects

Requirements

  • Google Gemini API key with embeddings capability
  • Pinecone account with an active vector index configured
  • X API credentials for automated posting

Service Value

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

Apps Used

AI Agent Vector Store Gemini Document Default Data Loader Text Splitter Output Parser LLM Chain Twitter/X

Details

Trigger Automatic trigger
Nodes 21
Apps 8
Views 221
Downloads 126

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 (21)

/

When clicking ‘Test workflow’

Manual Trigger

#1

AI Agent

Agent

#2

Pinecone Vector Store

Vector Store Pinecone

#3

Embeddings Google Gemini

Gemini Model

#4

Default Data Loader

Document Default Data Loader

#5

Recursive Character Text Splitter

Text Splitter Recursive Character Text Splitter

#6

Google Gemini Chat Model1

Gemini Model

#7

Google Gemini Chat Model2

Gemini Model

#8

Google Gemini Chat Model

Gemini Model

#9

Structured Output Parser

Output Parser Structured

#10

Sticky Note

Sticky Note

#11

Set Fields - URL and Webhook URL

Set

#12

Make a web request

HTTP Request

#13

Structured JSON Data Formatter

LLM Chain

#14

Webhook for structured data

HTTP Request

#15

Webhook for structured AI agent response

HTTP Request

#16

Sticky Note1

Sticky Note

#17

Sticky Note2

Sticky Note

#18

Sticky Note3

Sticky Note

#19

Information Extractor with Data Formatter

Information Extractor

#20

Sticky Note4

Sticky Note

#21

Similar Templates

Automated workflow

24 nodes

Automate - Update title and description -

26 nodes

Automate Microsoft Teams

17 nodes

Product: Idea to Shop Vector - Dimension 768

8 nodes

Connect Google Sheets with Slack

12 nodes

Give the AI ​​agent a long-term memory tool router

39 nodes