Project Name

GEO Insight Pro — AI Visibility Audit & Competitive Intelligence Platform

Problem Statement

As AI search (ChatGPT, Perplexity, Gemini) grows, brands need to understand:

Why they appear or don’t appear in AI answers

Why competitors rank higher

How to optimize for AI discovery

Whether AI answers about them are accurate

Traditional SEO tools don’t cover AI-generated answers. GEO Insight Pro fills this gap.

Solution Overview

A platform that audits brand visibility in AI-generated answers using live data from multiple AI models, web sources, and fact-checking. It provides actionable insights and fixes.

Core Features

  1. Multi-model AI Answer Comparison

Queries 3 models in parallel: GPT-4o, Gemini 2.5 Flash, Perplexity Sonar Pro

Side-by-side comparison of responses

Identifies differences in brand mentions, citations, and tone

Aggregates results for analysis

  1. Intelligent Query Generation

AI-generated query packs based on brand and topics

Toggle between manual and AI-generated queries

Intent-aware query tagging (course_intent, pricing_intent, career_intent, etc.)

  1. Web Evidence Collection

Scans top 10 web sources per query using Tavily API

Extracts content and analyzes brand/competitor mentions

Tracks intent terms (certificate, pricing, subscription, etc.)

Calculates citation overlap between AI and web sources

  1. Fact-checking System

Verifies 4 key facts: refund policy, certificates, accreditation, pricing model

Two-pass search: official domain + targeted keywords

Evidence snippets with source URLs

Classification: correct / unclear / wrong

  1. Competitor Analysis

Auto-detects competitors from AI answers and web sources

Threshold-based filtering (ai_mentions >= 1 OR web_mentions >= 2)

Entity visibility table (AI mentions vs web mentions)

Identifies why competitors win

  1. AI-powered Explanations

CrewAI agent (Gemini 2.5 Flash) generates explanations

“Why AI Missed This Brand” analysis

Evidence-linked fix suggestions with:

Placement (new page vs existing page)

Confidence level (High / Medium / Low)

Draft content (outline or FAQ format)

Evidence URLs

  1. Query Verdict System

One-line status per query:

"Brand missing everywhere"

"Brand appears in AI but weak in sources"

"Competitor dominates both"

"Brand visible in both AI and sources"

  1. Aggregate Scoring

Visibility score: Brand mention rate across queries

Accuracy score: Fact-check correctness (1 - hallucination_rate)

Clarity score: Evidence availability (1 - uncertainty_rate)

Competitor share: Per-competitor mention rates

Intent mention rates: Breakdown by intent type

  1. Modern Minimal UI

Monochrome design with subtle background patterns

Responsive layout

Expandable query details

Real-time metrics dashboard

Executive summary with insights

Technical Stack Backend

FastAPI — REST API

Python 3.x — Core language

Asyncio — Parallel processing

Pydantic — Data validation

AI / ML Services

OpenRouter API — Multi-model access (GPT-4o, Gemini 2.5 Flash, Perplexity Sonar Pro)

Google Gemini API — Direct fact-checking and reasoning

CrewAI — Agent orchestration for explanations

Perplexity API — Fallback for AI answers

Data Collection

Tavily API — Web search and content extraction

URL parsing and domain extraction

Frontend

Vanilla HTML / CSS / JavaScript

Minimal monochrome design

Responsive grid layouts

Innovation Points

Multi-model comparison: First tool to compare 3 major AI models side-by-side for brand visibility

Evidence-based fixes: Fix suggestions include draft content and evidence URLs

Threshold-based competitor filtering: Reduces noise with smart filtering (ai_mentions >= 1 OR web_mentions >= 2)

Intent-aware analysis: Multi-tag intent detection (course, pricing, career, best_of, comparison)

Real-time fact-checking: Verifies AI claims against official sources

Query verdict system: One-line status per query for quick insights

Use Cases

Educational platforms: Audit visibility for course-related queries

SaaS companies: Understand AI search presence

E-commerce brands: Optimize for AI discovery

Marketing teams: Competitive intelligence

SEO professionals: Bridge traditional SEO and AI search

Demo Highlights

Executive dashboard: Visibility, Accuracy, Clarity scores at a glance

Model comparison: Side-by-side view of GPT-4o, Gemini, Perplexity responses

Query analysis table: Expandable rows with full details

Competitor visibility: Entity comparison table

Fact-check module: Evidence snippets with classification

Fix generator: Actionable fixes with draft content

Technical Achievements

Parallel processing: Queries run concurrently for speed

Error handling: Type-safe operations prevent runtime errors

API integration: Multiple APIs orchestrated seamlessly

Responsive design: Works on desktop and mobile

Real-time data: No synthetic data; all results from live APIs

Future Enhancements

Competitor page scraping and analysis

Historical tracking of visibility trends

Automated fix implementation suggestions

Integration with CMS platforms

Team collaboration features

Built With

Share this project:

Updates