Publications · Case Studies

Operational Deployment, not claims

Each case follows one fixed template: the business problem, the AI solution, the evidence as it happened, then the findings, the chain completion audit and the governance assessment. Published cases are redacted.

02v0.4.0Chain completion 36%

Case 02 — AI-Native Inventory: Image-to-Unit Identification, Condition Capture and Per-Stay Reconciliation

Point the camera at the shelf. Four seconds later, the count exists — an AI capability embedded in Casa Admin's Add Inventory module, identifying items, quantity, condition and location from a single unstaged photo.

Enterprise AI Series · Run 3 August 2026 · 8 findings

A serviced property carries a large, heterogeneous item count per room. Checkout re-inventory took hours per room, done manually, inside the highest-pressure part of the operating cycle. An AI capability built into Casa Admin identifies items, quantity, condition and location from a single photo in about four seconds, reconciling against the existing inventory record and flagging duplicates before a human commits the result.

Read the case →

JD-01v0.5.0Chain completion 79%

Reservation Creation and Payment Reconciliation by an apeople

A documented run of Jeni De completing a full reservation chain across five systems on one natural-language instruction

Enterprise AI Series · Actor Jeni De (JD) · Run 30 July 2026 · 12 findings

A returning guest asked for a room, paid by UPI over WhatsApp, and a verified booking existed in three systems of record ten minutes later — with no human touching the reservation system, the finance ledger or the ticket queue.

Read the case →