IA Diffroom

IA Diffroom is a web app (with optional desktop companion for crawling) that treats information architecture as a versioned asset: sitemaps, navigation trees, content models, and internal link structures. It continuously snapshots your IA from sources you already have (XML sitemaps, CMS exports, analytics paths, and a crawler run) and turns changes into human-readable diffs: “12 pages moved,” “3 nav labels changed,” “orphaned nodes increased by 18%,” “new duplicate categories created.” Teams can open “IA pull requests” to propose changes, attach rationale, and route approvals to UX, SEO, and content owners before anything ships. This is not a magical auto-IA generator. It’s a practical governance layer that reduces accidental breakage during redesigns, CMS migrations, and ongoing content churn. The MVP focuses on a few high-signal checks (orphaning, depth bloat, broken trails, inconsistent labels, duplicate nodes) and produces a weekly IA health report executives can understand. AI is used sparingly: summarizing diffs, clustering similar labels, and suggesting review checklists—not inventing your taxonomy. Realistically, adoption will be strongest in mid-market orgs that already feel pain from “small” IA changes causing SEO drops, support tickets, and stakeholder chaos.