🏠 Wayfair Rugs Market Intelligence — AI Agent System

Built during the Wayfair AI Externship (2026) An end-to-end AI agent system that automates trend discovery, competitor monitoring, content strategy, and market intelligence for Wayfair's Rugs category team.

---

📌 Project Overview

This project is a suite of four connected AI agents built in n8n that transform hours of manual retail research into minutes of automated, structured intelligence. Each agent produces a formatted HTML report, and those reports feed into a unified Market Intelligence Dashboard — giving the Rugs category team one place to see trends, competitor pricing, whitespace opportunities, risk signals, and recommended actions.

Key Findings from Real Runs

• 📈 Rug market growing at 9.59% CAGR ($62.9B → $130.6B by 2033)
• 💰 $200–$400+ price tier is completely unserved by both Walmart and Amazon — a direct whitespace opportunity for Wayfair
• 🧺 Machine-washable 8×10 rugs at $120–$180 are the #1 volume sweet spot in the category

---

🤖 The Agents

Agent 1 — Trend Discovery Agent

Input: Category keyword + Amazon.com URLs Output: Full HTML trend report with micro-segments, market sizing, attribute analysis, and AI-generated moodboard images

• Scrapes Amazon product listings (titles, prices, ratings, reviews)
• Pulls social trend signals from Instagram, Pinterest, and blogs
• Identifies 2–3 emerging micro-segments via AI reasoning chain
• Generates visual moodboards using Hugging Face FLUX
• Runtime: 3–5 minutes

---

Agent 2 — Competitor Monitoring Agent

Input: Category keyword + Amazon and/or Walmart URLs Output: Side-by-side competitive intelligence report with pricing comparison, whitespace analysis, and supplier identification

• Scrapes Wayfair (auto-fetched), Amazon, and Walmart simultaneously
• Runs data through 6 sequential AI generators
• Identifies pricing gaps, key suppliers, and strategic recommendations
• Runtime: 8–12 minutes

---

Agent 3 — Market Intelligence Dashboard

Input: Completed P2 and P3 HTML reports saved in Google Drive Output: Unified HTML dashboard with 6 tabbed views

• One-click trigger in n8n — no text input required
• Auto-detects latest date folder in Google Drive
• Parses all AI-generated sections using custom HTML extractors (no Cheerio)
• 6 tabs: Executive Overview · Market & Trends · Competitive Intel · Opportunity Radar · Risk & Diagnostics · Action Center
• Runtime: under 2 minutes

---

Agent 4 — AI Insights & Content Agent

Input: Web form with category, focus area, and uploaded P2 + P3 HTML reports Output: Full content strategy report

• 6 data-backed content ideas (blog, social, video, email, Pinterest, buying guide)
• Ready-to-post captions for Instagram, Pinterest, Facebook, and TikTok
• 3 campaign concepts with taglines and competitive angles
• Email subject lines with predicted open rate tiers
• Competitive content angles vs. Amazon and Walmart
• Powered by OpenRouter AI
• Runtime: 2–3 minutes

---

🗂 Repo Structure

wayfair-rugs-market-intelligence/
│
├── workflows/                        # n8n workflow JSON files
│   ├── P2_Trend_Discovery_Agent.json
│   ├── P3_Competitor_Monitoring_Agent.json
│   ├── P5_Market_Intelligence_Dashboard.json
│   └── P4_Content_Strategy_Agent.json
│
├── sample-outputs/                   # Real HTML reports generated by the agents
│   ├── Area_Rug_Trend_Report.html
│   ├── Area_Rug_Competitor_Report.html
│   ├── Area_Rug_Dashboard.html
│   └── Area_Rug_Content_Strategy.html
│
├── dashboard-template/               # The base HTML template used by Agent 3
│   └── dashboard_template.html
│
├── presentation/                     # Final externship presentation
│   └── Wayfair_Final_Presentation.pptx
│
└── README.md

---

🛠 Tech Stack

Tool	Purpose
n8n	Workflow automation and agent orchestration
OpenRouter	AI model routing for analysis and content generation
Google Gemini	Image generation for moodboards
Google Drive API	Report storage and retrieval
Cheerio / Custom HTML parsers	Data extraction from scraped pages
JavaScript (n8n Code nodes)	Data transformation and HTML assembly

---

⚙️ Setup & Usage

Prerequisites

• n8n installed (self-hosted or cloud)
• OpenRouter API key → openrouter.ai
• Google Drive OAuth2 credentials configured in n8n
• Google Gemini API key (for Agent 1 moodboard generation)

Running an Agent

1. Import the workflow JSON into your n8n instance

   • Open n8n → Workflows → Import from file
   • Select the JSON from the /workflows folder

2. Add your credentials

   • OpenRouter: Settings → Credentials → Add OpenRouter Chat Model
   • Google Drive: Settings → Credentials → Add Google Drive OAuth2

3. Run Agent 1 (Trend Discovery)

   • Paste a category keyword (e.g., Area Rug) and Amazon URL(s) into the chat trigger
   • The agent runs and saves the HTML report to Google Drive

4. Run Agent 2 (Competitor Monitoring)

   • Paste a category keyword and Amazon + Walmart search URLs
   • Report saves automatically to Google Drive

5. Run Agent 3 (Dashboard)

   • Click Execute Workflow — no input needed
   • Dashboard HTML is available for download

6. Run Agent 4 (Content Strategy)

   • Open the n8n form URL in your browser
   • Fill in category, focus area, and upload P2 + P3 HTML files
   • Download the content strategy report

Google Drive Folder Structure Required for Agent 3

Wayfair Reports/
└── [Category]/           e.g., "area_rug"
    └── [YYYY-MM-DD]/     e.g., "2026-04-23"
        ├── P2_report.html
        └── P3_report.html
dashboard_template.html   ← must be in root of Wayfair Reports

---

📊 Sample Outputs

All four sample HTML reports are included in the /sample-outputs folder. Open any of them directly in a browser — no server required.

Report	Description
Area_Rug_Trend_Report.html	Full trend analysis with Boho Geometric, Modern Textured Neutrals, and Organic Sunlight-Inspired micro-segments
Area_Rug_Competitor_Report.html	Side-by-side Wayfair vs. Amazon vs. Walmart analysis with pricing gaps and supplier data
Area_Rug_Dashboard.html	Unified 6-tab intelligence dashboard
Area_Rug_Content_Strategy.html	Full content strategy with social captions, campaigns, and email subject lines

---

🔮 Future Improvements

• Connect all four agents into a single automated weekly pipeline
• Integrate real Wayfair SKU and sales data to validate trends against revenue
• Embed Chart.js visualizations so data renders as interactive charts
• Add a Wayfair brand voice profile to the Content Agent
• Expand to additional categories beyond the 4 supported rug types
• Publish the dashboard to a shared internal URL (no file download needed)
• Add error handling, automated retries, and monitoring for production use

---

👩‍💻 About

Built by Lauren Bolin as part of the Wayfair AI Externship (2026).

• 📧 labolin19@gmail.com
• 💼 LinkedIn ← update this link

---

📄 License

This project was built for educational purposes as part of an externship program. Workflows and sample outputs are shared for portfolio and learning purposes.