A product ofAurevus Innovations Inc. ↗·Governed operations, powered by AI

How you will know it worked.

Railspike does not publish savings figures. A demonstration shows you the mechanism on a reference portfolio; once deployed, the ledger measures what actually happened in yours — and your organisation decides what it was worth.

What it costs

C$4 per door per month. One rate, whole lifecycle.

A door is counted once, whether you are building it or running it — so a building moving into operations is not a re-quote. Buildings and projects organise your doors; they never add a second charge. Your whole team is included, because the point is that everyone works from the same record.

—

Authorized preview

Preview access is arranged directly. Public signup and purchasing are not open.

—

150 doors minimum

C$600 a month is the smallest subscription. Above that you pay for the doors you have, and the total moves as you add and retire them.

—

AI included, and stated

150,000 AI tokens per door each month, pooled across the workspace — roughly 27 questions per door. You can see what you have used.

Plus applicable taxes. Above 2,500 doors, talk to us.

Why there is no number on this page

We cannot know your baseline, and we will not estimate it.

Every vendor in this category can produce a return-on-investment figure. Producing one requires assuming your loaded costs, your current recovery rate, how completely your team adopts the system, and what you would have caught anyway. We know none of those, and a confident number built on four assumptions about your business is not evidence — it is a sales artifact that your CFO will correctly discount.

01

What a demonstration shows

The mechanism, running on a reference portfolio you are free to inspect. You see how a lapsing certificate is connected to the work and the money that depend on it, and the arithmetic that produced the figure. It is our data, and we say so.

02

What it does not show

Your numbers. Those require your records, and your records require a deployment. Any figure about your business before that point would be a guess wearing a decimal point.

03

Who decides what it was worth

Your operational owners, HR and Finance. Railspike is designed to calculate and evidence, never to approve its own value claim.

If you already run a tight ship

Then the value is not what we find. It is what stops depending on a person.

A well-run operator may have very little sitting exposed, because somebody is holding it together — a controller who checks the certificate before the payment run, a superintendent who remembers which sub is lapsing. That is vigilance, and it is real work being done well.

→

Vigilance does not scale

It holds at four hundred units and at one project. The same standard across four thousand units and six projects is a different job, and it is usually the reason the next hire is needed.

→

Vigilance does not survive turnover

The check that exists because one person knows to do it leaves when they do. A control that runs whether or not anyone remembers is a different kind of asset.

→

Vigilance is not evidence

“We caught it” is not the same as being able to show a lender, an auditor or an insurer when it was caught, by whom, and against which rule.

We would rather find nothing in your registers than tell you we found something you were already handling. If a deployment surfaces little, that is a finding about your operation worth having — and the case for Railspike becomes continuity and evidence rather than recovery.
The problem being solved

The exposure lives between the systems, so none of them can price it.

A vertically integrated operator carries the same building through development, construction, turnover and stabilised operations — and the same subcontractor rough-ins the building, then services it for twenty years. Every system in the stack is correct on its own data. Each is missing the half that turns a fact into a liability.

One vendor, four systems
The safety register knows a clearance expires — and watches nothing downstream of it.
↓
The maintenance system knows which jobs sit in that vendor’s scope — and cannot see a certificate.
↓
The draw package knows what their work is worth this period — in a different application, owned by a different department.
↓
Accounts payable approved their last claim, because the approver could see none of it.
What the product does

Mechanism, not forecast.

These are descriptions of behaviour you can verify in a demonstration. If the system does not do them, that is a defect — not a missed projection.

✓

It reads the registers

Clearances, contracted obligations, progress claims and draw lines — from the source files those systems already produce.

✓

It states the exposure

What depends on one lapsing document, across both sides of the business, as a single figure with a date attached.

✓

It holds the consequence

Work orders held, the payment gate closed, the draw refusing to submit incomplete — each with the rule that fired and its effective date.

✓

It closes only on evidence

A hold is released by a document whose dates actually cover the work, checked arithmetically, or by a named person. Never by a timer.

The value ledger

Six levels, and only the last two are money.

Most software reports activity and calls it value. The ledger is designed to keep them apart, so the monetary headline moves only when the evidence supports something more concrete than a saved minute. It measures a deployment that is running; it forecasts nothing before one exists.

Level 1–2Activity and coverage

What ran, and what portion of eligible scope it observed. Coverage matters more than clicks — and if coverage is low, the ledger says so rather than flattering the total.

Level 3–4Output and capacity

Findings raised, obligations escalated, evidence assembled — and the handling time your managers confirm was removed. Hours, reported as hours.

Level 5–6Validated and realised

Cash recovered, an expense prevented, a position absorbed by released capacity. Each ratified by a named authority in your organisation, with the evidence attached.

Recognition discipline

Not every hour becomes a dollar.

A credible ledger reports released capacity even when it has never touched the budget, and reports outcomes that went the wrong way.

Evidence stateHow it is treated
EstimatedModelled from disclosed assumptions. Excluded from the monetary headline entirely until validated.
Validated capacityA functional owner confirms the work changed. Reported in hours, separately from cash.
Avoided staffingHR and Finance confirm growth was absorbed without an otherwise required position.
RealisedCash recovered, or an expense prevented, reconciled to a source-system record.
Reversed or disputedRemoved from current totals — without deleting the original record.
And the instrument is not exempt. A measurement layer can be miscalibrated like anything else. Every figure it produces carries the baseline it was measured against, the assumptions applied and the person who ratified it — so a number you disagree with can be traced, challenged and corrected rather than defended. Where the evidence does not support a figure, the ledger is designed to report the lower state, not the more flattering one.
No double counting. Where released capacity later enables an avoided hire, the avoided-hire value supersedes the monetary value of those hours. The hours stay visible as evidence; they are not added twice. Loosening any recognition factor requires the approval path your organisation configured, and leaves a receipt.

See the mechanism before you decide anything.

The most useful thing we can show you is not a slide. It is a question answered across four systems that answer it separately — on a reference portfolio, with the arithmetic shown, which you are welcome to check line by line. What it finds in your records is a question a deployment answers, not a sales call.