Information

"It's like documentation, but less informative"

🧭 Table of Contents

💾 Overview

This page covers information to help users understand Serca beyond the basic level. It also covers service behavior, architecture concepts, and how the pipeline works end-to-end.

🛠️ Architecture Overview

At a high level, Serca starts from a set of seed URLs, discovers related pages/subdomains, scrapes and extracts content, runs an AI pipeline to generate metadata, and finally updates the database for searching.

Serca pipeline (concept)

This is a simplified diagram. It shows the data flow and the AI stage components at a glance.

Seed URLsWeb ScraperExtract SubdomainsSubdomains → ScraperAI PipelineUpdate DatabaseOCRYOLOv5BLIPImage Features

Scraping + discovery

Starting from seed URLs, Serca discovers additional pages and subdomains, then scrapes content and metadata for analysis.

AI enrichment

The AI pipeline extracts text (OCR), detects objects (YOLO), generates captions (BLIP), and computes image features for stronger matching.

💳 Pricing

Serca will offer a free tier and paid tiers. The free tier is designed to be useful and safe; higher tiers enable larger query limits and additional capabilities.

Exact pricing may change as infrastructure costs and indexing scale. The goal is predictable billing and clear limits.

📦 API

Right now, Serca’s API is for backend/internal use while the platform stabilizes. A public API is planned once rate limiting, authentication, and documentation are ready.

Coming soon

  • API keys + usage dashboard
  • Rate limiting + tier-based quotas
  • Public docs + examples

SQL is magic...