Automotive Retail Intelligence Platform
One vehicle record. Every department smarter.
RooftopOS keeps one record for every VIN, and every application and department writes to it. Start with the workflow that solves today’s problem; the record is what makes the next one worth more than the first.

Product availability today
- AutoCurbLive
- AutoLabelsLive
- AutoFilmPilot
- Initial founding design-partner deployment
The operational problem
The vehicle is the only object every department shares.
Dealership software is bought a department at a time. Each system is correct about its own slice of a car and blind to the rest, so the same vehicle ends up described several separate ways before it is ever sold.
The cost lands at the boundaries. Work gets repeated because the evidence for it was lost at a handoff, a disclosure is built from data nobody verified, and an appraisal starts from nothing on a car the store has already owned once. None of those systems is wrong. Nothing in the stack is organised around the one thing they all have in common.

Vehicle Acquisition
Acquire with confidence.
Off-street submissions, service-lane vehicles and in-store trades run through one pipeline — decode, condition, appraisal, manager approval, customer offer, check request. Every adjustment is attributable to a person and a time, and the condition evidence behind the number stays attached to the vehicle rather than to the conversation.
It reads decoded identity, trim and factory equipment off the shared record, along with any inspection history already held against that VIN. It writes back the acquisition channel, the appraisal and the approval behind it, the cost basis, and the check-request trail — so the reason for a number survives the person who made it.

Vehicle Identity and Merchandising
Publish from verified vehicle data.
Identity and factory equipment are decoded and verified from the VIN, the customer-facing record is built from that verified data rather than retyped from a trim guess, and the advertised price is checked against the store's own live site before anything publishes. Every version is retained.
It reads VIN identity, decoded equipment and recall status, plus the condition and acquisition context already on the record. It writes back the window sticker, the Buyers Guide, the dealership addendum and the customer-facing Vehicle Passport — each one generated from the record rather than assembled by hand into a template.

Sales and Service Video
Turn vehicle video into action.
A salesperson walkaround or a technician's inspection is recorded against the specific vehicle, the next step sits inside the video rather than in a follow-up call, and the customer's answer is retained on the record instead of in one person's phone.
It reads the vehicle the video belongs to, together with open recommendations and previously declined work. It writes back timestamped findings with the evidence still attached, and the customer's approve, question or decline decision — which is what the next department reads instead of asking again.

Continuous vehicle history
Nothing is thrown away at a handoff.
Entries written by three different applications, across the life of one vehicle, on one record. Read in order, that sequence is the whole argument for a shared record: no department starts from nothing.
Each application writes to the same vehicle record. Automated cross-application handoffs remain in development.

The expansion model
The second application is worth more than the first.
A dealership never has to buy a platform to get value from one. Defensibility is a mechanism here rather than a claim: every workflow adds structured vehicle history to the same record in the ordinary course of business, so the next application opens against a record that is already populated.
The shape below is a direction, and it says so in its own field. The model is real; the scale is not a forecast and is not presented as one.

Founding design partner
Designed against a working store.
The product is designed against a working store rather than against an assumption about one — a used-car director, a service drive and an accounting office that all have to live with it on the days nobody has time for software. What that buys is a workflow that survives an ordinary week. What it does not buy is proof at scale, and this page claims none.
Measurement is part of the arrangement rather than an afterthought: how the work is done today is captured before anything changes, and the evidence is reviewed against it. No result from that work appears here, and none will until the methodology, the comparison and the sample size can be published alongside it.

Data ownership and trust
Commitments a dealership can hold RooftopOS to.
How access is configured, where one rooftop’s boundary sits, and the terms under which a dealership exports its own records are written down rather than described. The Trust Center states each of them, alongside the controls that are in place and the ones that are not.
Nothing on this page describes a certification the company does not hold. What it describes are commitments, and the agreement each of them is written into.
Prove RooftopOS in one rooftop.
This is an evaluation, not a trial. It is designed so that at day sixty both sides are looking at the same numbers and the decision is obvious to whoever has to sign it.
It needs an executive sponsor at the store, a named day-to-day owner, and access to the inventory feed and to the website that publishes advertised prices. Scope and commercial terms are confirmed in the scoping conversation, in writing, before anything starts.

One record. Every application. Every department.
Start with the application that solves the problem in front of the store today. RooftopOS is what makes the next one worth more than the first.
RooftopOS is taking on dealership partners and is open to investor conversations. Request an investor briefing
