From Flyer to Feed: A Vision for Hyperlocal Event Discovery

The Problem

Every city has two event calendars: the official one and the real one.

The official one lives on Eventbrite, Facebook, and venue websites. It's searchable, structured, and incomplete. The real one is stapled to telephone poles, taped to coffee shop windows, and pinned to community bulletin boards. It's comprehensive, chaotic, and unsearchable.

The best events—the neighborhood block party, the DIY show at the bar, the pop-up market in the church parking lot—live on paper. They exist as flyers designed in Canva, printed at Staples, and distributed by hand.

This content is invisible to the web. It's invisible to AI assistants. And it's invisible to anyone who didn't happen to walk past the right telephone pole.

The Idea

What if anyone could snap a photo of a flyer and turn it into a structured, searchable, syndicatable event?

RSVP-System could offer a simple pipeline:

  1. Capture — Upload a photo of a flyer (mobile, email, SMS, web form)
  2. Extract — AI reads the image and pulls out event details: title, date, time, venue, cost, description
  3. Review — A human reviews the extracted data against the original flyer, approves or edits
  4. Publish — The event goes live with proper Schema.org markup, iCal export, and API availability

The flyer stays attached as provenance. The community gets credit for contributing. The event becomes discoverable everywhere.

Why This Matters

For discovery: When someone asks "what's happening this weekend," the answer should include the weird, wonderful, grassroots stuff—not just ticketed concerts and corporate gatherings.

For preservation: Flyers are ephemeral. Rain destroys them. Events pass. A photo archive with structured metadata becomes a historical record of community life.

For accessibility: Not everyone walks the same streets. A centralized, searchable hub democratizes access to hyperlocal information.

For AI readiness: Structured event data with Schema.org markup is exactly what AI assistants need to answer natural language queries. A well-maintained local hub becomes the authoritative source.

The Centaur Workflow

This isn't about replacing human judgment—it's about eliminating tedious transcription.

AI handles: Reading text from images, parsing dates, extracting venue names, suggesting categories.

Humans handle: Verifying accuracy, resolving ambiguity ("Saturday" without a year), matching venues to existing records, deciding what gets published.

The flyer submitter contributes local knowledge. The AI contributes speed. The editor contributes quality control. Everyone wins.

Open Questions

  • Intake friction: How simple can we make submission? Snap-and-send via SMS? Forwarded email attachments? A PWA with camera access?
  • Venue matching: How do we handle variant spellings and informal names? Fuzzy matching? Community-maintained aliases?
  • Trust and moderation: Open submissions invite spam. What's the right balance between openness and quality?
  • Attribution: How do we credit contributors without creating privacy concerns?
  • Flyer rights: Who owns a flyer image? Is republishing it fair use? Do we need permission?

Next Steps

  1. Prototype the AI extraction prompt using Claude's vision capabilities
  2. Define the Drupal content type fields needed for flyer-sourced events
  3. Build a minimal review queue UI
  4. Test with a small batch of real Philly flyers
  5. Evaluate accuracy and editorial burden

The Bigger Picture

RSVP-System is about making reservation and event management accessible to organizations that can't afford enterprise solutions. This feature extends that mission: making event discovery accessible to communities that don't have a marketing budget.

Every city deserves a comprehensive, structured, open calendar of what's happening. The data exists—it's just trapped on paper. Let's free it.


This is a build note from Gluebox LLC's work on RSVP-System, an open-source Drupal platform for reservation and event management. Follow along at [link] or subscribe to the RSS feed.