Ledger reads your financial statements against the IFRS disclosure checklist, decides what applies, and returns each verdict with the page it came from — ready to paste into your working paper.
Functional currency disclosed. → p. 142
Cash & equivalents reconciliation incomplete. → p. 168
Entity holds no insurance contracts. → profile
An IFRS disclosure checklist runs into thousands of conditional clauses, cross-referenced against notes that span hundreds of pages. The standards revise every year. The work is exacting, the margin for omission is wide, and almost all of it lands on the senior reviewer.
It is the cost of every other working paper losing the benefit of the doubt. When the review fails an obvious item, the signing partner has no choice but to revisit the file end-to-end. The economics of audit hinge on what gets caught on the first pass.
Ledger keeps a working paper feel — every output is something a reviewer could have produced by hand, and is meant to be signed off the same way.
Drop the PDF of the financial statements. Ledger parses the structure — primary statements, notes, schedules — and indexes every paragraph against the page it lives on.
Before answering anything, Ledger reads the entity: its sector, instruments, segments, listing status. This profile is what makes the difference between "absent" and "not applicable".
For every applicable check item, Ledger searches the document, returns Y, N, or N/A, cites the page, and writes the reasoning a reviewer would write themselves.
Step through each verdict with the page rendered alongside, override what you choose, then export back to the xlsx checklist your team already uses.
Each one is a deliberate response to a way the current generation of audit AI tools tends to fail.
Most AI review tools answer in two columns: found, or not found. The not-found column is where the false positives live. Ledger reasons about applicability first — it distinguishes absent because non-existent from missing where required. N/A is treated as a first-class verdict, not a fallback for confusion.
Every verdict carries the page or pages it came from, with the relevant zone rendered visually inside the review interface. There is no black box to defend: the reviewer sees the evidence before signing.
The output is an xlsx in the disclosure-checklist format the team already uses. No proprietary portal to onboard onto. The reviewer receives what they would have produced by hand — verified, justified, and ready in minutes instead of weeks.
Decisions you can defend, infrastructure your IT director will approve, evidence your file partner can sign.
Deploy inside your own infrastructure if needed. In on-premise mode, source documents never leave your perimeter.
Every verdict, every override, every export — recorded with a timestamp and a user. Designed in the spirit of ISA 230 (Audit Documentation).
An admin account governs the team. Reviewers see their own engagements only.
IFRS standards revise yearly. Each run is frozen against the checklist version it used, so historic engagements remain reproducible.
No verdict ships without a page reference. Every conclusion points back to the document.
The model proposes. The reviewer decides, overrides, and signs. By design.
Two plans. Pay per run as a solo practitioner, or a monthly subscription with degressive top-up for firms that review at scale.
One run is one complete audit of an annual report: the PDF is parsed, every applicable IFRS disclosure requirement is evaluated, and each verdict comes back with its evidence images. Billing is metered on the tokens the model processes: 12 EUR per million input tokens and 60 EUR per million output tokens. In practice, a typical report reviewed against a ~2,000-item checklist comes to around 100 EUR; the exact price varies with the size of the report and the number of check items. No subscription, no commitment — you pay for each run you launch.
Individual suits auditors who work alone on their engagements. Enterprise becomes the right fit as soon as several auditors collaborate on the same files — multiple accounts, the ISA 230-aligned review workflow, custom exports — and it is also the more economical option at a sustained monthly volume.
The 2,000 EUR monthly plan includes roughly 10 runs. Beyond that, usage is metered per token at rates well below the Individual plan — and the rate itself decreases as your monthly volume grows.
| Monthly token spend | Input / 1M | Output / 1M | Approx. per run |
|---|---|---|---|
| 0 – 2,000 EUR | 6.00 EUR | 30.00 EUR | ~50 EUR |
| 2,000 – 5,000 EUR | 5.10 EUR | 25.50 EUR | ~43 EUR |
| 5,000 – 12,000 EUR | 4.50 EUR | 22.50 EUR | ~38 EUR |
| Above 12,000 EUR | 3.90 EUR | 19.50 EUR | ~33 EUR |
Rates are per million tokens processed by the model, input and output metered separately, and cover the full service: the retrieval pipeline, PDF parsing, evidence-image rendering and infrastructure. The approximate per-run price is indicative for a typical 2,000-item report; it varies with the size of the report and the checklist. This degressive scale reflects our own marginal cost, which decreases with volume.
Reviewing financial statements against a comprehensive IFRS disclosure checklist (~2,000 conditional items) typically takes a senior auditor between 10 and 20 hours of focused work. Depending on market and seniority, fully-loaded auditor time ranges from 30 to 70 EUR per hour. That places the human cost of one full disclosure review somewhere between 300 and 1,400 EUR per report.
The same review runs at ~100 EUR on the Individual plan — the return is positive from the very first audit, before any of the qualitative gains below. On Enterprise, the degressive tiers bring the effective per-run cost down further as your volume grows.
But hourly cost is only part of the picture. The work an auditor doesn't do when our pipeline handles the first pass is the work where mistakes are most likely and least valuable. The real ROI compounds across four dimensions.
A single reviewer applies the same standards across 200 engagements that the same human auditor would interpret slightly differently from one Friday afternoon to the next. Every entity gets the same depth of scrutiny.
Every verdict ships with a cited source page, a visual highlight, and a reproducible reasoning chain. Working papers are review-ready and ISA 230-aligned by construction, not by post-hoc cleanup.
Your senior auditors stop spending their best hours filling cells in a spreadsheet. They review exceptions, exercise judgment, advise clients — the work that justifies their seniority and protects your firm's margin.
Manual disclosure review fails in two ways: items missed entirely (omission), and N/A cases misread as non-compliance (false positives). The first creates audit risk for your firm. The second wastes review-meeting time and erodes client confidence. Both are dramatically reduced when applicability is checked systematically.
At ~100 EUR per run — and less at volume — the cost sits well below the human equivalent in nearly every market, before considering the qualitative gains above.
We wanted to keep the barrier to entry low for solo practitioners. You only pay when you actually run an audit. If your workflow requires team collaboration or if your monthly volume grows significantly, Enterprise unlocks additional features and better rates.
One full audit of an annual report against a chosen IFRS checklist. The pipeline parses the PDF, retrieves relevant evidence, evaluates each disclosure requirement, and produces verdicts with page-level citations. Re-running the same report (for example after a checklist update) counts as a new run.
Additional runs are billed via the top-up system with degressive pricing. The more your firm consumes, the lower the effective cost per run — the degressive tiers are designed to reward volume without forcing you into a bigger commitment upfront.
Yes. You can start on Individual and move to Enterprise at any time — just get in touch.
Send us the kind of engagement you'd normally run by hand. We'll show you what Ledger returns.