Skip to main content
Toggle navigation
Work

Strong evidence

PE Scanner

Evidence-backed ownership transparency with model-assisted research and human publication review.

Tier
Strong evidence
Pillar
Production systems
Visibility
Sanitized
Status
in-development
Date
Jul 2026

Case boundary

A native iOS application and ownership service that separates storefronts, brands, operators, corporate parents, franchisors, public companies, and private-equity sponsors through sourced, time-aware relationships.

Claim boundary

Development classifications are unpublished previews, not a public coverage claim.

PE Scanner is a consumer ownership-transparency system. Chicago bagel storefronts are the first vertical used to validate discovery, entity resolution, evidence extraction, review, and publication.

System

  • The iOS 26 SwiftUI app provides list and map browsing, ownership filters, brand grouping, evidence detail, saved locations, and research-stage previews.
  • FastAPI and Postgres model locations, brands, organizations, identifiers, aliases, time-aware relationships, assertions, evidence documents, evidence spans, reviews, publications, and source runs.
  • Discovery uses public business data and place sources; DocPull acquires documents; structured model output proposes entities and relationships.

Control

  • Model extraction creates review candidates only. An authenticated human must approve assertions before publication.
  • Source URLs and exact supporting excerpts are validated against stored content.
  • Missing parent evidence always produces Unknown, and Independent requires affirmative evidence.

Proof

  • The graph supports cycle detection, relationship validity dates, supersession, classification precedence, and publication-readiness gates.
  • Current development classifications are explicitly unpublished research previews and are not represented as public coverage.
  • Backend tests and iOS unit and UI tests cover the native and service surfaces.