An AI-powered research tool that analyzes AI search citations, evaluates source coverage, and identifies topics where stronger authoritative information may be needed.
The RAG Analysis Engine was built to explore how AI search systems retrieve information, select supporting sources, and construct answers.
Users enter a topic or niche idea, and the engine:
- Searches for relevant sources
- Reviews the citations returned
- Evaluates source coverage
- Identifies potential information gaps
- Generates practical content recommendations
- Provides links to the supporting sources
The project began as an experiment in finding “information deserts”: subjects where AI systems may have limited, fragmented, or insufficiently authoritative material available.
RAG Analysis Engine
https://rag-analysis-engine.vercel.app/
The repository also includes GEO Pulse, a separate experimental engine available at:
https://rag-analysis-engine.vercel.app/geopulse
GEO Pulse analyzes a business website in relation to a target-market keyword.
- AI-powered search analysis
- Live citation and source discovery
- Information-gap detection
- Opportunity summaries
- Actionable content recommendations
- Clickable supporting sources
- Expandable technical JSON output
- Error handling and input validation
- Responsive user interface
The engine can help users explore:
- underserved content topics
- niche business ideas
- emerging product categories
- local-market information gaps
- weakly documented industry questions
- opportunities for authoritative educational content
- Next.js
- React
- TypeScript
- Tailwind CSS
- Firecrawl API
- OpenAI API
- Vercel
app/
├── api/
│ └── void-check/
│ └── route.ts
├── geopulse/
│ └── page.tsx
├── test/
│ └── page.tsx
└── page.tsx
The application separates the user interface, API routes, source retrieval, and analysis logic so the engine can be expanded into additional research and recommendation tools.
Clone the repository:
git clone https://github.com/peachsites/rag-analysis-engine.gitOpen the project:
cd rag-analysis-engineInstall the dependencies:
npm installCreate a local environment file:
cp .env.example .env.localAdd the required environment variables:
FIRECRAWL_API_KEY=
OPENAI_API_KEY=
Start the development server:
npm run devOpen:
http://localhost:3000
The engine identifies potential opportunities based on the sources returned during each analysis. Results should be treated as research guidance rather than a guarantee of search rankings, traffic, citations, or commercial success.
Source availability and model responses may change over time.
Future improvements may include:
- saved searches
- historical opportunity tracking
- competitor comparisons
- downloadable reports
- multi-keyword research
- business visibility monitoring
- user accounts
- subscription features
Built by Randa
© 2026 Randa. All rights reserved. This repository is provided for demonstration and portfolio purposes.