PURSER
PATENT PENDING · U.S. PROV. APP. 64/114,845

The officer who settles the fleet's accounts.

Purser reads the settlement statements a distribution network already produces — and finds the losses sitting in plain sight inside them.

26/26
consecutive settlement statements reconciled to the cent
$7,500
manufacturer-borne stale loss found on one route, one quarter — previously invisible on both sides
56 days
between when the loss was detectable and when it was booked
64/114,845
provisional patent application, filed July 2026
What it is

An intelligence layer, not a new system

Every week, a distribution network's settlement statements record everything that happened on a route — sales, returns, stale write-offs, chargebacks, a per-SKU inventory reconciliation. Purser reads that same data and validates it, tracks velocity on forced-allocation programs, times guarantee windows against stale allowances, and flags what a route owner or a manufacturer would otherwise only discover after the loss is already booked.

No new data capture. No workflow change for the field. An intelligence layer on statements already being generated today.

Ingest

Sign-preserving extraction of every settlement section, as filed.

Reconcile

Every statement validated against its own accounting identity before anything downstream runs.

Detect

Velocity and depletion trajectories on forced-allocation programs, flagged by week four.

Attribute

Each risk positioned against its guarantee window and allowance — a dollar figure, a party, a deadline.

Where it came from

Built from inside the network

Purser began as one operator's own settlement pipeline — years of running an active distribution route, reconciling weekly statements by hand, and finding a systemic blind spot in the process. The platform is the tool that was built to close it, tested first on real money before being offered to anyone else.

Get in touch

Talk to us about a pilot

Purser is running today. If you operate or oversee a distribution network and want to see what it finds in your own data, we'd like to talk.