build

AI Wedding Guest Concierge — MVP Build

"Brainstorm and build an AI wedding guest concierge for Adventure Weddings. Couples design their wedding, then an AI agent handles all guest coordination."

4 branches explored 0

AI Wedding Guest Concierge — Morning Report

Session: 12 min active (research + build) Mode: Hybrid (research → build) Status: MVP demo complete


Executive Summary

Built a working demo of an AI wedding guest concierge for Adventure Weddings. The concept: couples design their destination wedding, then an AI agent handles all guest coordination — the most stressful part of any destination wedding.

The research phase confirmed a wide-open market opportunity. No existing wedding platform offers an AI concierge for guests. Every competitor (The Knot, Zola, Joy) focuses AI on helping the couple plan. The closest analogy is hotel concierge AI (Runnr.ai, HeyGuest) — which proves the pattern works in hospitality.


What Was Built

FilePurpose
.claude/skills/wedding-concierge/SKILL.mdCLI skill for processing guest messages
apps/adventure-weddings/weddings/sample/wedding.yamlSample Queenstown helicopter wedding
apps/adventure-weddings/weddings/sample/guests/*.json4 sample guests (AU, US, NZ, IN)
apps/adventure-weddings/src/pages/guest/[token].astroGuest chat page with demo AI
apps/adventure-weddings/src/pages/couple/dashboard.astroCouple dashboard with stats
apps/adventure-weddings/src/pages/concierge.astroProduct landing page

Routes Added

  • /concierge — Product landing page
  • /guest/demo — Demo chat experience
  • /couple/dashboard — Guest management dashboard
  • Build: 9 pages in 722ms, all clean

Market Opportunity

  • No direct competitor in AI concierge for wedding guests
  • Closest: Daisy Chat (early-stage SMS chatbot, reactive only)
  • Proven pattern: Hotel concierge AI (Runnr.ai, HeyGuest, ConciergeBot)
  • Core value: “Design your wedding. We’ll handle your guests.”

Key Pain Points Solved

  1. Couples become unwilling travel agents for 30+ guests
  2. Guests don’t know the area, timezone, visa rules, what to pack
  3. Weather-dependent plans require real-time communication
  4. Group coordination is manual and tedious
  5. Dietary/accessibility collection is usually a spreadsheet nightmare

3-Phase Architecture

Phase 1 (now): Claude Code CLI as backend. File-based storage. Zero cost. Phase 2 (next): Astro + Cloudflare Functions. Web chat with Claude API streaming. ~$5 per 100-guest wedding. Phase 3 (future): Multi-wedding platform. Self-serve setup. Custom branding.


Individual Findings

Finding: 01-market-gap

No existing wedding platform offers AI concierge for guests. Every competitor focuses on the couple. Wide-open market opportunity.

Finding: 02-hotel-ai-validation

Hotel concierge AI (Runnr.ai, HeyGuest) proves the pattern works. Same concept, different vertical.

Finding: 03-mvp-demo

Working demo with guest chat, couple dashboard, and product landing page. 9 pages, 722ms build time.

Finding: 04-architecture-plan

3-phase rollout from CLI-based MVP to multi-wedding platform. Phase 1 costs $0.