Developer & DevOps

[1/3 - Anomaly detection] [1/2 - Classification algorithm classification] Upload the dataset to Qdrant (set...

25 nodes 263 145 Automatic trigger
Download

Workflow Description

Workflow that detects and classifies anomalies using classification algorithms, uploads processed datasets to Qdrant for storage and retrieval, leveraging Google Cloud Storage for centralized data management and processing pipelines.

How it works

  1. 1.Trigger workflow manually and retrieve dataset from Google Cloud Storage
  2. 2.Process data and apply classification algorithm to categorize records
  3. 3.Execute anomaly detection model on classified dataset
  4. 4.Filter results based on defined criteria using conditional logic
  5. 5.Send processed data to Qdrant via HTTP request
  6. 6.Log detected anomalies and classification outcomes

Use cases

  • Identify suspicious transactions and anomalous user behavior in financial systems
  • Automatically classify and organize large datasets for improved retrieval and analytics
  • Monitor data quality and ensure compliance across big data infrastructure

Requirements

  • Active Google Cloud Storage account with full access permissions
  • Configured Qdrant instance ready for vector ingestion and similarity search

Service Value

Ready-made workflow template for automation delivery and service execution.

Apps Used

Google Cloud Storage

Details

Trigger Automatic trigger
Nodes 25
Apps 1
Views 263
Downloads 145

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 (25)

/

When clicking ‘Test workflow’

Manual Trigger

#1

Google Cloud Storage

Google Cloud Storage

#2

Get fields for Qdrant

Set

#3

Qdrant cluster variables

Set

#4

Embed crop image

HTTP Request

#5

Create Qdrant Collection

HTTP Request

#6

Check Qdrant Collection Existence

HTTP Request

#7

Batches in the API's format

Set

#8

Batch Upload to Qdrant

HTTP Request

#9

Split in batches, generate uuids for Qdrant points

Code

#10

If collection exists

If

#11

Sticky Note

Sticky Note

#12

Payload index on crop_name

HTTP Request

#13

Sticky Note1

Sticky Note

#14

Sticky Note2

Sticky Note

#15

Sticky Note3

Sticky Note

#16

Sticky Note4

Sticky Note

#17

Sticky Note5

Sticky Note

#18

Sticky Note6

Sticky Note

#19

Sticky Note7

Sticky Note

#20

Sticky Note9

Sticky Note

#21

Sticky Note11

Sticky Note

#22

Filtering out tomato to test anomalies

Filter

#23

Sticky Note8

Sticky Note

#24

Sticky Note10

Sticky Note

#25