A core ledger shadow for Canadian banks

Banking changes.
Accounting confidence.

A new core release. A product rule change. A migration. Know what changed in the accounts—and why—before making the release decision.

In research & development · Seeking early design conversations

THE SHADOW LEDGER CONCEPT
01 / OBSERVED

What the core posted

02 / EXPECTED

What the rules require

A difference is a starting point.

Trace it to the event, rule, or mapping.

A second view of your books.

Evidence for your next change.

Decisions owned by your team.

01 / THE APPROACH

Your core tells you what happened.
A shadow helps explain why.

SpiritLedger is being designed to give Technology and Finance a shared view of accounting outcomes, alongside the institution’s existing core.

01

Replay the history

Reconstruct account activity from authorized exports, opening balances, and versioned chart mappings. Establish the input boundary before comparing results.

A declared starting point
02

Compare the outcomes

Compare baseline and candidate outputs against independently reviewed expectations, where business events and product rules are available.

An independent expectation
03

Follow the evidence

Connect differences to source records, calculation dates, and configuration versions. Assemble reproducible evidence for the people making the decision.

A reviewable explanation

02 / A SMALL DIFFERENCE. A REAL QUESTION.

The journals balance.
Is the interest right?

In our synthetic deposit case, a candidate applies a rate change one day late. Every journal still balances. Two customers receive less interest than the stated expectation.

The useful question goes beyond the total: which day diverged, which rate applied, and what evidence supports the explanation?

Interactive illustration from our research specifications. These are synthetic, arithmetic-checked amounts—not output from a running SpiritLedger engine.

CASE / CAD DEPOSITS

October interest review

Synthetic
Synthetic interest comparison in Canadian dollars
Interest paid · CADExpectedCandidate
Account A51.5050.30
Account B24.2023.70
Total75.7074.00
CAD 1.70 underpayment

First divergence: October 16 · Rate effective date

The candidate starts the new rate on October 17 instead of October 16. Account A is short CAD 1.20; account B is short CAD 0.50.

03 / BUILT AROUND THE DECISION

One history.
Several reasons to look closer.

01

Core releases & migrations

Investigate whether a candidate preserves the intended accounting behaviour before the scoped change is approved.

Engineering + Finance
02

Product rules & chart changes

Explore interest, fees, corrections, and classification changes with their effective dates and historical meaning intact.

Product + Financial Control
03

Replay & recovery

Plan for another reviewer to reproduce the case from a consistent ledger, chart, configuration, and evidence set.

Operations + Technology

04 / START WITH ONE CHANGE

Bring a question
from your books.

We’re looking for conversations with Canadian banking teams working through deposit-platform changes. Help shape a focused validation workflow around a decision your team actually needs to make.

Start a conversation

Opens a public GitHub inquiry. Share only non-sensitive project details.

A proposed first pilot

  • One CAD deposit product
  • A history crossing month-end
  • One authorized batch export
  • Agreed accounting expectations
  • Reviewed findings and replay evidence

Scope and acceptance would be agreed with your Technology and Finance owners. Product readiness and data access determine what can be validated.

A FEW PRACTICAL QUESTIONS

Before we talk.

Is SpiritLedger available today?

SpiritLedger is in research and specification. The ledger engine and application stack are still being evaluated. This site introduces the proposed product and invites early design conversations; it does not describe an available production service.

Would this replace our core?

The proposed initial scope is a shadow alongside your existing core: replay, reconciliation, and validation. The institution’s core remains authoritative. Payment execution is outside the proposed initial scope.

What data would a pilot need?

Opening balances, posted journals, chart and mapping history, and completeness evidence are the starting point. Independent interest validation also requires business events, rates, calendars, and agreed product rules. Journals alone cannot independently prove calculation correctness.

Who should join the first discussion?

Your core-platform lead, accounting or Chart of Accounts owner, and testing or reconciliation lead. Together we can identify an upcoming change, the decision owner, and a bounded question worth investigating.