Conso 360.

AI accounting software

Less typing, the same accounting control.

Conso 360 uses AI to read supplier documents, classify them and propose the accounting. Your team reviews and approves the result, while deterministic rules handle VAT, duplicates and period locks.

Built for: accountants, finance teams, outsourced accounting providers and SMEs.

Where AI helps

  • Reading supplier invoices and receipts
  • Recognising the document type
  • Suggesting accounts, VAT and expense categories
  • Proposing bank transaction classifications
  • Mapping columns in spreadsheet imports

The data entry problem

Most accounting time still goes into copying information from documents.

Supplier invoices arrive as PDFs, photos and e-invoices in different layouts. Someone has to read each one, type the details, pick the account and VAT treatment and check it has not been booked before.

Repetitive capture

Supplier, dates, amounts and lines are re-typed from documents that already contain them.

Inconsistent coding

The same supplier is coded to different accounts by different people.

Missed duplicates

The same invoice arrives by email and by post, and is entered twice.

Month-end backlog

Documents pile up until the close, when there is least time to review them.

How a supplier invoice is processed

From inbox to posted invoice, with a person in the loop.

A practical example of what changes for an accountant processing supplier invoices.

  1. 01

    Receive

    Invoices arrive through a connected mailbox, an email forwarding address or upload. Structured e-invoices are read directly.

  2. 02

    Extract

    OCR and AI extract supplier, dates, amounts, VAT and lines, and recognise whether the document is an invoice, receipt or pro forma.

  3. 03

    Suggest

    Accounts and VAT are proposed from your rules first, then from previous postings, with AI only as a fallback.

  4. 04

    Review

    The document waits in a review queue with confidence indicators and duplicate warnings. Your team corrects and approves it.

  5. 05

    Learn

    Corrections are fed back into the classification rules, so the next invoice from that supplier needs less work.

AI-assisted and deterministic

Clear about which parts are AI and which are rules.

AI is used where documents are unstructured. Accounting decisions that need to be predictable are handled by rules you can see and change.

Document extraction (AI-assisted)

Text recognition and AI extraction read supplier invoices and receipts in different layouts and languages.

Document classification (AI-assisted)

Documents are classified by type before extraction, and unclear cases are flagged for review.

Account and VAT suggestions (rules first)

Your classification rules and posting history are applied before any AI suggestion, and every suggestion can be changed.

Bank transaction classification

Matching rules classify bank lines; AI proposes a classification with a confidence score only when no rule applies.

Duplicate detection (deterministic)

Possible duplicate invoices are detected and shown before approval.

Optional auto-approval

You can allow documents below an amount limit that pass confidence, risk and plausibility checks to be approved automatically. It is off unless you switch it on.

Product interface

A review queue, not a black box.

Each document shows what was extracted, what is suggested and how confident the system is, so reviewers spend time on the exceptions.

Document queue · needs review

Illustrative example
SupplierAmountSuggested accountConfidenceCheck
Office Supplies Ltd € 248.00 Office expenses High Ready to approve
Cloud Hosting OÜ € 1,120.00 IT services High Possible duplicate
Fuel station receipt € 63.40 Vehicle costs Low Check VAT

Nothing is posted until it is approved by a person or by thresholds your team has set.

Simplified illustration with fictional companies and figures. It is not a screenshot of customer data.

What AI does not do

AI-assisted accounting, not autonomous accounting.

Conso 360 does not replace your accountants. Extraction can be wrong, especially on poor scans or unusual layouts, which is why every result is reviewable.

  • Accounting treatment, VAT judgement on unusual transactions and period-end estimates remain with qualified people.
  • AI suggestions can be overridden, and corrections are kept as feedback rather than silently ignored.
  • Auto-approval applies only where you enable it and only above the thresholds you set.
  • AI usage is logged per feature, so you can see where it is used and what it costs.

Controls

The checks stay in place whatever the source of the data.

Documents processed with AI go through the same accounting controls as manual entries.

  • Review before posting

    Documents move from extracted to needs review to approved, and only approved documents become purchase invoices.

  • Period and posting locks

    Postings into closed periods are blocked, regardless of how the document arrived.

  • Exception handling

    Failed extractions, possible duplicates and documents waiting for review are flagged instead of being skipped.

  • Traceability

    The original document stays attached to the invoice, and changes to invoices are versioned.

Frequently asked questions

What does AI accounting software actually automate?

In Conso 360, AI reads and classifies supplier documents, extracts the invoice data and suggests accounts and VAT where your rules do not already decide. It also helps classify bank lines and map spreadsheet columns during imports. Posting, VAT logic and period controls stay rule-based.

Will AI post entries without anyone checking?

Not by default. Documents wait in a review queue until someone approves them. You can enable auto-approval for documents above confidence and risk thresholds you choose.

Which documents can be processed?

Supplier invoices, receipts, pro forma and advance invoices in PDF or image form, plus structured e-invoices such as UBL XML, which are read without AI.

How accurate is the extraction?

It depends on document quality and layout. That is why each result shows confidence indicators and why the review step exists. We prefer to show you on your own documents in a demo rather than quote a percentage.

Does it learn from our corrections?

Yes. Corrections to accounts and categories are stored as feedback and classification rules, so later documents from the same supplier are suggested more consistently.

Does this replace our accountant?

No. It removes repetitive data entry so that accountants can spend their time on review, judgement and advice.

Related guides

Try it on your own supplier invoices.

In a demo, we process a few of your documents and show what is extracted, what is suggested and what your team still decides.