Skip to content

Overview

Guide

The Knowledge Base interface is organized in four tabs:

TabDescription
OverviewAI-generated and editable workspace information blocks
StorageImported documents and files
PipelinesAutomated data import pipelines
EmbeddingsView of generated vectors (visible only for Super Admin role)

  1. Import documents — upload files or configure pipelines to feed the base
  2. The system processes — documents are converted into vector embeddings
  3. Agents use it — AI agents query the base to respond with context
  4. Dashboards monitor — track usage and processing on the Home Screen

On the Overview tab you can automatically generate the product information blocks with AI, using the magic button (✨ Sparkles).

When generating, Prodgy retrieves the content already stored in the active Knowledge Base (via semantic search) and uses it as context. This way, the result reflects the product’s real knowledge — not just the name, description, and website provided.


FormatExtensions
SpreadsheetsCSV, XLSX, XLS
DocumentsPDF, DOC, DOCX
TextTXT, MD, Markdown
Structured dataJSON, XML

The Knowledge Base uses Server-Sent Events (SSE) to keep data automatically updated. Changes such as uploads, deletions, and status updates are reflected in the interface without needing to reload the page.


Knowledge Base operations consume plan credits:

  • Document upload — credits per import operation
  • Embedding generation — credits per vector processing
  • Storage — credits per MB stored