← Back to owner console

Help / Glossary

RAG Prep Cloud operating guide

A simple owner/customer explanation layer for ETL, RAG, crawler credits, vector search, templates, payment monitoring, traffic, and launch operations.

RAG

Retrieval-Augmented Generation: the pattern of retrieving relevant company knowledge and giving it to an AI system so answers are grounded in real business data.

ETL

Extract, Transform, Load: collect data from files, websites, APIs, or systems; clean/normalize it; then load it into a searchable destination.

Crawler

A controlled website collection workflow that discovers pages, fetches content, respects default safety policy, and exports clean content for RAG/ETL.

Vector search

Semantic search using embeddings so users can find meaning-similar content, not just exact keyword matches.

Clean Markdown

A cleaned text output format that removes most webpage clutter and gives customers readable, RAG-friendly content.

Credits

Metered units that control paid crawler and advanced-feature usage such as pages crawled, JS rendering, extraction, and document/RAG processing.

Extraction template

A saved schema or instruction set for pulling structured fields from crawled pages or uploaded documents.

Owner console

The master operations page for traffic, users, Stripe/payment evidence, crawler jobs, credits, launch readiness, and system health.

How customers use the product

  • •Upload files or run an authorized crawl.
  • •Convert raw content into clean text, chunks, metadata, and embeddings.
  • •Search through the dashboard or API.
  • •Export crawler outputs as JSON, CSV, or clean Markdown.

How the owner monitors the product

  • •Open the owner console for total/daily visits, users, billing customers, crawler jobs, credits, and Stripe evidence.
  • •Open ops pages for admin readiness, crawler ops, auth, and payment rollout gates.
  • •Use Stripe for payment records and webhook evidence.
  • •Use Vercel for deployment status and logs.

Why RAG Prep Cloud matters

  • •Companies avoid weeks of custom data-prep work.
  • •Internal documents and websites become searchable AI-ready knowledge.
  • •Teams can ship grounded AI features faster than competitors still manually cleaning data.
  • •The crawler, ETL, and RAG layers are metered so usage can map to revenue.