AI Systems
Paaing: Recursive People Also Asked Research SaaS, 64+ Questions Per Keyword, AI Articles Instant
SEO software development that recursively extracts 64+ People Also Asked questions per keyword across 200+ Google domains, plus one-click AI article generation.
Django
Material UI
OpenAI GPT-4o
PostgreSQL
Python
React
200+
Google Country Domains
10x
More PAA Data Than Standard SEO Tools
64+
Questions Per Depth-3 Search
The Problem
Your Strongest SEO Content Opportunities Are Hidden at PAA Depth 2 and 3
SEO software development has largely ignored one of the richest signals for understanding how real users think about a topic: People Also Asked boxes. Standard tools only scratch the surface, making simple HTTP requests to Google search result pages, missing the dynamically loaded PAA boxes entirely, and never expanding the tree recursively. At depth 2 or 3, the question graph grows exponentially, and that deeper data is where the real content strategy gold lives. Content teams were manually clicking expand on Google PAA boxes, copying questions into spreadsheets, and spending hours writing answers and articles. There was no tool that combined recursive PAA extraction with AI content generation in a single workflow.
Our Solution
Recursive PAA Expansion Across 200+ Countries with AI Content Generation
Paaing is built around Google's undocumented internal callback endpoint, bypassing the standard SERP HTML page entirely and hitting the same JSON endpoint Google's own UI calls to lazily load PAA boxes. Country-specific query context tokens and a large static parameter block that mimics a Chrome browser session are pre-computed and injected per request. The scraper runs entirely async, all Level N parent keywords are processed simultaneously, making depth-3 searches fast even for complex, broad topics. A rotating proxy pool handles antiban. The graph builder converts the flat question list into a visually organised interactive tree. Three AI generation modes, snippet, short article, long article, are powered by the user's own API keys (OpenAI or Anthropic), keeping operating costs at zero for the platform while giving users full model flexibility.
Architecture
Django 5 + DRF API. Django Channels WebSockets for real-time scraping progress. Celery + Redis for async task queue. aiohttp with rotating proxy pool for concurrent PAA scraping. BeautifulSoup HTML parsing. PostgreSQL with composite index on (keyword, country, language, depth, pid). OpenAI and Anthropic API integration with user-supplied keys. NowPayments crypto billing. AppSumo/Dealify code redemption system. React + Redux + Material UI frontend.
The Result
64+ Questions Per Keyword, Real-Time PAA Research and AI Article Generation
Paaing gives agencies a capability no standard keyword tool can match, depth-3 recursive PAA expansion that surfaces 64+ questions per keyword, including the long-tail search intentions most competitors never discover. One seed keyword generates a complete content calendar's worth of research in seconds. Content teams using Paaing consistently produce 10x more SEO-optimised content from the same time investment. Instead of manually clicking expand in Google and copying questions into spreadsheets, they get a structured question tree in real time, ready for one-click AI article generation. The 200+ country and Google domain combinations gave global SEO agencies fresh, country-specific question data for every target market, in one platform. If your business needs a similar SEO or content research tool built, from recursive data extraction to AI-powered article generation, our team can scope it the same way.
Want scalable architecture and measurable results like this?
Book an Architecture Audit
Build Something Like This
// free scoping call · senior engineers · no commitment
Project scope, timelines, and architecture options in one 45-minute session.
More Work