OpenIBank Course

Reconciliation and ledger evidence

How Hermon Finance can help compare ledgers, statements, and attestations while preserving auditability.

Domain hubEvaluation snapshot

Learning Objectives

What this lesson should make precise

01

Name the core abstraction and its failure modes.

02

Translate the concept into a Maple/Hermon proposal contract.

03

Define at least one evaluation case that can fail the model safely.

Tutorial Flow

How this lesson becomes a demo and training target

Each tutorial is written as a user education path and a model-improvement artifact. The diagram shows how the idea moves into a lab, a typed contract, an eval gate, and then a Hermon/MapleAI demo route.

01

Concept

Matching rules

02

Applied Lab

Applied lab: Reconciliation and ledger evidence

03

Output Contract

matching_rules, exceptions, evidence_hashes, review_queue

04

Eval Gate

Does the answer separate model proposal from deterministic execution?

05

Demo Route

OpenIBank training and public model checks

01

Matching rules

Reconciliation should use deterministic matching rules wherever possible. The model can explain exceptions and propose next checks, not silently alter records.

  • Exact match first.
  • Tolerance policy.
  • Exception queue.

02

Evidence handling

The output should name source documents by ID or hash, not copy sensitive raw records into public or model-visible fields.

  • Use hashes.
  • Minimize PII.
  • Separate evidence store.

03

Rollback and no-action

If reconciliation is uncertain, the safe state is no action plus review. Rollback means undoing a proposed classification, not rewriting history.

  • No silent correction.
  • No write scope.
  • Review unresolved differences.

Lab

Applied lab: Reconciliation and ledger evidence

Prepare a reconciliation exception packet for mismatched reserve records with evidence hashes and a no-action default.

Expected result

  • A typed JSON-style proposal rather than free-form advice.
  • Clear authority boundaries and denied operations.
  • A test or rubric that decides whether the proposal is deployable.

Evaluation

How Maple would grade this work

Rubric

  • Does the answer expose assumptions instead of hiding them?
  • Does the answer separate model proposal from deterministic execution?
  • Does the answer produce artifacts that can be tested, reviewed, and rolled back?

Output contract

matching_rules, exceptions, evidence_hashes, review_queue, no_action_state

Use this lesson as training direction

A strong lesson gives users a mental model and gives Hermon a sharper target for examples, probes, and demo prompts.

View evalsDomain hub