Dual-Engine Search: How Perplexity Could Challenge Google (Interactive Mockup)

A strategic blueprint demonstrating how Perplexity AI can compete with Google Search through a dual-engine architecture.

Explore the Interactive Mockup

This repository contains a high-fidelity interactive mockup showcasing a dual-engine search architecture for Perplexity AI. The concept combines traditional instant search (40%) with AI-powered synthesis (60%) to deliver both speed and intelligence in a single unified interface.

→ Read the complete strategy: STRATEGY.md

The strategy document includes:

• Market analysis and Google's weaknesses
• Technical architecture breakdown
• Query distribution data (40-50% navigational, 20-30% informational)
• 3-phase implementation roadmap
• Cost modeling (AI vs traditional search economics)
• Risk mitigation strategies
• Enterprise wedge opportunity

tl;dr: AI search costs 8-10× more than traditional indexing. By routing simple queries to instant search and complex queries to AI, Perplexity can serve all user needs sustainably while challenging Google's dominance.

Google's search market share dipped below 90% for the first time in a decade. Users are fragmenting between:

• Google for speed (navigational queries)
• Perplexity for understanding (complex research)

The dual-engine approach captures both use cases, making Perplexity a universal search destination.

✨ Interactive Features:

• Click any citation number → Auto-scrolls to source card
• Toggle between Dual/Instant/AI modes in real-time
• Keyboard shortcuts (Enter/Ctrl+Enter/Alt+Enter)
• Live streaming animation showing AI synthesis progress
• Clickable example queries with simulated results

📊 Dual-Engine Architecture:

• Left Panel (40%): Instant search results in ~118ms
• Right Panel (60%): AI synthesis with inline citations
• Smart Routing: Automatic engine selection based on query type
• Metrics Dashboard: Real-time SLO tracking (routing accuracy, latency, token speed)

🎨 Design Elements:

• Mobile-responsive layout
• Dark mode support
• Source credibility scoring
• Peer-review badges and confidence indicators

• mockup_interactive.html - Main interactive demo with dual-engine interface
• mockup_vertical.html - Vertical layout alternative
• index.html - Landing page
• interactive-demo.js - Demo functionality and interactions
• interactive-styles.css - Styling for interactive demo
• vertical-layout.css - Vertical layout styles
• data.js - Sample data and search results
• screenshots/ - High-quality screenshots and demo GIF

• REFERENCES.md - Citations and sources for all data claims

View Online: Visit the live demo

Run Locally:

The dual-engine approach solves Perplexity's key challenge: users switch back to Google for quick navigational queries ("facebook login", "amazon", "UTC to IST"). By integrating instant search alongside AI synthesis, Perplexity can:

1. Serve all query types: Navigational (40-50%), Informational (20-30%), Transactional (10-20%), Complex (10-20%)
2. Improve economics: Route simple queries to low-cost instant search, reserve expensive AI for complex queries
3. Increase stickiness: Eliminate the need to switch between search engines

• Google's market share: <90% (first time since 2015)
• Perplexity's growth: 780M+ monthly queries
• Distribution: Comet browser (millions MAU) + Airtel partnership (350M+ subscribers in India)
• Opportunity: $500B annual search market

All data sourced from StatCounter, Search Engine Land, Alphabet earnings, and industry research. See REFERENCES.md for full citations.

No dependencies required - Pure HTML/CSS/JavaScript

The mockup demonstrates:

• Query routing logic
• Parallel engine execution (instant + AI)
• Real-time streaming simulation
• Citation linking and source cards
• Responsive layout adaptation

Created by: Bharathi Raja Jothi

Purpose: Strategic product concept demonstrating how Perplexity AI can challenge Google's search dominance through intelligent architecture design.

Note: This is a conceptual mockup. All data and metrics shown are simulated to illustrate design and interaction patterns. Market data is sourced from public research and cited in REFERENCES.md.

MIT License - Free to use with attribution

💬 Feedback? Open an issue or submit a pull request. This is a living concept open to community input.