One grounded engine. Two ways in.
Attera reads source documents and pins every figure to its page. Accounting firms buy that as a productized Operating Report on their clients' statements. Financial institutions come to us through an introduction for the grounded data underneath their own reporting. The extraction is the same. What sits on top of it, and how you buy it, is not.
For the ESG team doing this by hand, at scale.
You are at a bank or an asset manager, and the reporting you owe starts with figures buried in documents: statements, filings, supplier data, in more than one language. Someone re-types them, checks them, and hopes they hold when an auditor asks where a number came from. That is the job the engine does, at whatever scale you have.
The grounded layer, not the finished disclosure.
We turn your source documents into reconciled, cited data and hand it over with an evidence pack. Your team makes the judgment calls and writes the disclosure, with our ESG lead scoping and reviewing the work. We do not author your CSRD report, and we will say so in the first meeting rather than the tenth.
- Grounded extraction on the datapoints you choose, CSRD/ESRS, every figure cited to its source page
- Reconciliation against your own figures, with gaps and mismatches flagged rather than guessed
- A tamper-evident evidence pack with a SHA-256 manifest, verifiable offline
- Processing on our own hardware in Belgium, with no external AI on the document path
- Scoped as a paid pilot on your documents, priced per engagement talk to us
There is no plan to sign up for here and no price on the page, because the scope is different every time. If this is the shape of your problem, email contact@attera.io and we will tell you honestly whether the engine helps.
For the firm retyping client statements by hand.
You run an accounting firm or fiduciary with a book of independent clients: insurance agents, telecom resellers, brokers, freelancers. Every month their commission and income statements arrive in a different layout, and someone keys them into a reconciliation by hand. Attera reads those statements and returns the Operating Report, every line cited to its source. One account, a dossier per client.
Branded per client. Scoped per client.
A Subscription account holds up to 15 active client dossiers, and the Firm tier goes to 60 across offices. Each export carries your firm's logo or the client's. Evidence packs are scoped per client, so when you hand one client's pack to their auditor, none of your other clients' documents are inside it.
- Up to 15 active client dossiers on Subscription, 60 on Firm
- Branded PDF exports · your logo or the client's
- Per-client traceability and audit log
- Commission-statement reconciliation pipeline fiduciary use case
- Annual Income & Reconciliation XLSX export
- External auditor read-only seats included
Worked example: commission reconciliation.
An illustrative scenario, in the shape of our Swiss back-office partnership (see the case): one mid-sized broker, twelve insurance carriers, eight hundred commission statements a month, each PDF with a different layout. The same grounded pipeline produces the Annual Income and Reconciliation report on top.
- Ingests multi-carrier statement PDFs, extracts commission lines with page citations
- Extracts every commission line with page citations, ready for ledger reconciliation ledger import on roadmap
- Flags discrepancies for human review rather than guessing a number
- Exports an Annual Income & Reconciliation XLSX, ready for the broker's auditor
- Per-client workspace isolation: one client's evidence pack contains only that client's source files and audit log entries
One firm, one subscription, a dossier per client.
You already hold the client relationship and you already do this work by hand, which is the whole reason we start here. One subscription covers your dossiers under a single account, branded per client. The monthly close is the job we are trying to give you hours back on, and it is the only thing we are asking you to buy.
The same engine, pointed at your documents.
The two motions above are the productized paths. Underneath, the engine reads any document with structured figures. The rest is scoped per engagement, on request, and labelled here by how far along it is.
Other document families
Invoices, operational reports and back-office categories, run through the same grounded pattern as the reports above. See the Swiss case →
EU Taxonomy Article 8 KPIs
Regulation 2020/852 alignment KPIs extracted with the same page-level citations.
iXBRL / ESEF filing package
A structurally valid iXBRL preview and ESEF package. The issuer-extension taxonomy layer is on the roadmap to filing-grade.
Deeper ESRS coverage
48 indicators auto-extracted today against 1,144 catalogued, with extraction depth being widened.
SFDR datapoint extraction
The engine can be pointed at an SFDR datapoint set on a scoped, paid engagement. Not built off-the-shelf, and we do not author the report.
Have a document family that isn't here? Bring a real one to a demo and we will tell you honestly whether the engine helps. Book a demo →