Reality

Building the intelligence infrastructure for social life.

The thesis: Loneliness epidemic meets AI phase change. LLMs can now parse unstructured event data that was previously impossible to automate. Start with the best events API for a single city, then layer on ranking, personal context, and coordination.

Current phase: Source registry research — mapping every data source for events in Philadelphia before writing a line of scraper code.

GitHub repo • Vision doc

Research Progress

Cat 1 Cat 4 Cat 2+3 Cat 5 Synthesis

Completed Research

Category 1 Philly Venue Landscape

Venues & Locations — OSM Overpass, OpenDataPhilly, Google Places, Yelp, Free Library RSS

4,882 venues via OSM 3,400+ city facilities 7 sources tested

Category 4 City & Government Data

OpenDataPhilly catalog scan, business licenses, event permits, Free Library RSS deep dive, PPR programming

72 library events (8 feeds) 3,332 event food licenses 0 permit datasets published

Upcoming Research

Categories 2+3 Organizations & Event Platforms

Meetup API, Eventbrite, Reddit, AllEvents.in, Luma, Partiful, Google Events — where do groups and platforms list events?

Category 5 Universities & Institutions

Penn, Temple, Drexel event calendars. Philadelphia Museum of Art, Barnes Foundation. Insider access to Penn data.

Synthesis Master Findings & Build Sequence

Coverage matrix across all categories, gap analysis, NYC transferability, recommended scraper build sequence.

Architecture

Layer 1 Events API ← current focus Layer 2 AI Ranking Engine Layer 3 Personal Context / CRM Layer 4 Action & Coordination Layer 5 Community Layer

Each layer gets its own brainstorm → plan → build cycle. We're in Layer 1 research: mapping every data source before writing scrapers.