Product Overview

What We're Building

An event planning platform that helps people organize events without the usual complexity. Core features: AI-generated event plans, block-based planning system, vendor discovery through web search, and a social portal where guests can RSVP and share photos.

Problem

When planning a wedding and other events, we ran into:

• Didn't know what kind of venue to use or what we needed at all
• Don't want to set everything up from scratch each time
• Finding vendors is time-consuming and you don't know who's reliable
• Event photos end up scattered across different phones and lost

Solution Approach

Core Planning System

Events are built from "blocks" - modular components like venue, catering, entertainment. Each block contains tasks, notes, and budget items. Users can organize blocks in a draft area for brainstorming, then create multiple plans to compare approaches. One plan becomes the active plan for execution.

This gives flexibility without overwhelming users with a blank canvas.

AI Integration

LLMs can generate decent event plans from basic inputs (event type, budget, guest count). We use this to give users a starting point they can modify. Also useful for vendor search and budget suggestions.

The AI is additive - the platform works fine without it, but it speeds things up.

Vendor Discovery (Phase 1: Search)

Start with web search integration - AI searches for vendors and presents options. We cache results to build our own database over time. This avoids the complexity of building a marketplace while still solving the vendor discovery problem.

Phase 2 (future): Full marketplace where vendors can join, get reviews, handle bookings.

Guest Portal

Each event gets a public-facing website where guests can:

• RSVP with dietary restrictions, plus-ones, etc.
• Upload and view photos
• See event details and updates
• Participate in discussions

This solves the photo collection problem and gives events a social element.

Target Users

Starting with consumers planning personal events:

• Weddings (complex, high-value, people care a lot)
• Birthday parties
• Family reunions
• Small community events

Later expansion to event planning businesses (they need team features, client management).

Technical Architecture

• Frontend: React SPA for planning interface, separate portal for guests
• Backend: AdonisJS API, PostgreSQL for data, Redis for sessions
• Storage: Cloudflare R2 for photos/documents
• AI: OpenAI API for plan generation and chat
• Vendor Search: Web search APIs (Google, Bing, or specialized)
• Deployment: Cloudflare Pages (frontend), DigitalOcean (backend)

Key Technical Decisions

Why JSON snapshots for versioning?

• Simple to implement
• Storage is cheap (few KB per snapshot)
• Avoids complex event sourcing
• Easy rollback to any version

Why separate draft area and multiple plans?

• Users need to experiment without breaking their plan
• Clear mental model (draft = brainstorming, plans = structured approaches)
• Enables "what-if" scenarios and plan comparison

Why blocks instead of tasks?

• Events naturally decompose into vendors/areas (venue, food, music)
• Blocks can have different schemas without database migrations
• Templates become collections of blocks

Why web search for vendors instead of marketplace?

• Marketplace is complex (onboarding, quality control, payments)
• Search gives immediate value without vendor buy-in
• We can build database from search results
• Natural progression: search → directory → marketplace

Revenue Model

Subscription Tiers

• Free: Individual users, basic features
• Individual (TBD pricing): Personal event planning with enhanced features
• Team/Business (TBD pricing): Multiple users, team collaboration features

Add-ons

• AI credits (pay-as-you-go for heavy users)
• Photo storage (for events with lots of guest uploads)
• Premium templates
• Custom domains

Future Revenue (Vendor Marketplace)

Once we have vendor marketplace:

• Commission on bookings (3-5%)
• Lead generation fees
• Featured listings
• Vendor subscriptions for advanced features

MVP Scope

Must Have

• Event creation with basic blocks (5-7 types)
• Draft area and multiple plan creation
• Task management within blocks with basic collaboration (invite users, assign tasks)
• Budget tracking
• Basic AI plan generation
• Vendor search via web APIs
• Simple guest portal with RSVP and photos
• User authentication and workspaces
• Task-level comments (nice to have)

Potential Future Features

Some features we're considering based on initial thinking (will be refined based on user feedback and business needs):

• Chat-based AI iteration
• Vendor database building from searches
• Advanced collaborative planning (real-time editing, permissions, full comment system)
• Advanced portal features (discussions, live updates)
• Mobile apps

Future Business Features

As we expand to the B2B market, we'll need business-focused features. Potential areas include:

• Team management and permissions
• Client relationship management
• Project management tools specific to event planning
• Integration with business systems
• Enterprise features

The exact features will depend on user research and market demand as we scale.

Vendor Strategy Evolution

Phase 1: Search (MVP)

• Integrate web search API
• AI processes results and recommends vendors
• Show contact info, websites, basic details
• Cache results to build knowledge base

Phase 2: Directory

• Build structured vendor database from search cache
• Allow users to save preferred vendors
• Basic vendor profiles (auto-generated from web data)
• User can add notes about vendors

Phase 3: Marketplace

• Vendors can claim and manage profiles
• Reviews and ratings system
• Direct booking through platform
• Payment processing
• Lead distribution

Risks and Mitigations

Risk: AI costs spiral with usage Mitigation: Credit-based system, usage limits, optimize prompts

Risk: Photo storage gets expensive Mitigation: Compression, limits, optional Google Drive integration

Risk: Web search API costs/limits Mitigation: Cache aggressively, multiple API providers, rate limiting

Risk: Vendor data quality from web search Mitigation: AI validation, user feedback loop, manual curation for popular vendors

Risk: No vendor buy-in for future marketplace Mitigation: Build value first (send them leads), prove demand, easy onboarding

Development Priorities

1. Get basic event + block system working
2. Add AI plan generation (simple version)
3. Integrate vendor search API
4. Build minimal guest portal
5. Launch MVP and get feedback
6. Build vendor cache/database from search usage
7. Iterate toward marketplace based on demand

Open Questions

• Which search API? Google vs Bing vs specialized (Yelp, etc)?
• How much vendor info to cache vs fetch live?
• Should we notify vendors when users search for them? (for marketplace validation)
• How to handle vendor search in different geographic areas?
• Build or buy review system when we get to marketplace phase?