Comarch ERP Optima Invoice Scanning After KSeF

Comarch ERP Optima invoice scanning is available through Comarch OCR, so the short answer is yes: Optima can capture supplier invoices from scans, photos, and PDFs and bring that data into the workflow. But after KSeF, the real question changed. You no longer need OCR for every purchase invoice, only for the documents that still arrive outside the structured KSeF channel.

That shift is driven by the 2026 rollout. According to the Ministry of Finance's guidance on the scope of mandatory KSeF, receiving invoices through KSeF became mandatory on February 1, 2026, most taxpayers had to issue invoices in KSeF from April 1, 2026, and suppliers without a Polish establishment are excluded from mandatory KSeF issuance. For many Polish businesses, that creates a practical split: domestic invoices increasingly arrive as structured KSeF data, while foreign-origin invoices and other out-of-scope documents still land as PDFs, email attachments, or scans that need capture before posting into Comarch ERP Optima.

That is why Comarch Optima invoice OCR after KSeF is no longer mainly about proving that OCR exists inside the ERP. The live decision point is whether native OCR is enough for the non-KSeF invoices you still receive, especially foreign-supplier PDFs, scanned documents, multilingual layouts, and line-item-heavy files. This article is a workflow decision guide for that narrower but still important problem, not a setup manual.

What Comarch OCR Actually Does Inside Optima

Comarch OCR is the native capture layer that helps you scan invoices into Comarch ERP Optima without retyping the basic purchase invoice fields by hand. In practice, it reads a scanned supplier invoice, extracts key data, and uses that data to prefill the purchase invoice workflow inside Optima. For a finance team already posting cost invoices in Optima, the appeal is straightforward: keep document capture close to the ERP instead of building a separate intake process first.

From a buyer's point of view, Comarch Optima invoice OCR is less about "does OCR exist?" and more about what it can actually populate reliably. Comarch positions the service to work with common invoice file types such as PDF, JPG, TIFF, and PNG, and to recognize data points that matter in bookkeeping and AP work, including:

• supplier identifiers
• VAT amounts and VAT rates
• net and gross totals
• currency
• invoice line items

That last point matters. If your team needs more than header-level posting, line-item capture can be the difference between basic registration and usable cost analysis, coding, or split posting. Native OCR is most useful when your incoming invoices are reasonably consistent in layout and your goal is to move from document to purchase invoice in Optima with as little manual correction as possible.

The commercial signals also matter when you assess fit. Comarch presents OCR as a cloud service, not just a one-time local feature. It publicly points to 87% accuracy, offers a 50-page demo, and gives buyers a visible Comarch OCR pricing benchmark through package pricing, including PLN 130 for 250 pages. That tells you two things early: first, you can test it on a small real sample before committing; second, OCR volume has a direct per-page cost, so capture quality and exception rates affect the economics.

For many Polish SMEs and accounting firms, that native model is a reasonable starting point. If most supplier invoices follow familiar patterns and you mainly want purchase invoices to land faster inside Optima, Comarch OCR can cover a meaningful part of that job. But native OCR exists is not the same as every invoice import scenario is solved. The real test is whether your remaining invoice mix still behaves like the documents native OCR handles best.

Which Invoices Still Sit Outside KSeF and Need Capture

KSeF reduces manual handling for structured domestic B2B invoices, but it does not eliminate your capture workload. The remaining work sits in the documents that still arrive outside that channel: foreign supplier invoices, emailed PDF invoices, scanned paper documents, and other purchase records that never show up as structured KSeF data. If you need a refresher on where the mandate applies and where it does not, our Poland KSeF rollout timeline and scope article breaks that out in more detail.

For teams using Optima, the operational problem is not just scanning volume. It is the mismatch between unstructured supplier documents and the fields Optima needs for booking, control, and approval. A foreign vendor PDF may contain the right facts, but they are still trapped in layout-specific text, multiple languages, different date formats, line-item tables, or tax descriptions that need review before posting. The same is true for invoices forwarded from shared mailboxes, attachments sent directly by suppliers, and scanned documents from branches or accountants' clients.

That is why Comarch ERP Optima invoice capture still matters after KSeF. Your team still has to identify the supplier, invoice number, dates, amounts, currency, VAT treatment, and often line-level detail before the document becomes usable inside the accounting workflow. Even when the invoice is perfectly legible to a human, it can still create AP work because Optima needs structured values, not just a readable PDF or image.

This is especially visible in foreign supplier invoice import into Comarch ERP Optima. The challenge is rarely "can we open the file?" It is whether the data can be extracted consistently enough for coding, validation, and approval without someone retyping or checking every field.

In practice, the post-KSeF AP workflow now splits into two lanes:

• Domestic B2B invoices move through KSeF and then into Optima
• Foreign or non-KSeF PDFs go through OCR or extraction, then review, then import into Optima

That second lane is where native OCR still matters, and where a more flexible extraction workflow can become the better fit.

---

Where Native OCR Stops Being Enough for Non-KSeF Documents

Native OCR inside Comarch ERP Optima still has a real advantage: it keeps capture close to the accounting workflow you already use. If your remaining non-KSeF invoices are mostly consistent supplier PDFs with predictable header fields and limited line-level complexity, staying inside the ERP can be the lowest-friction option.

The fit starts to change when the documents left outside KSeF are the messy ones. That usually means foreign-supplier invoices, multilingual layouts, scanned attachments, dense item tables, credit notes mixed with invoices, and batches where each supplier formats data differently. In those cases, the question is no longer just whether Optima can read text from a file. The harder question is whether your team can turn that file into clean, standardized data that is ready for review and import.

That distinction matters. Capturing text is not the same as producing import-ready data. A finance team may need one row per invoice, or one row per line item, with dates normalized, VAT fields handled consistently, credit notes shown as negatives, and output mapped into a repeatable structure before it ever reaches Optima. Native OCR can help with document intake, but when the invoice mix is irregular, teams often need more control over field rules, output layout, and exception handling than a basic in-ERP capture step provides.

This is where an extraction-first workflow becomes useful. Imagine a German supplier PDF that includes German VAT labels, multiple line items, and a credit note in the same batch. OCR may read the text, but finance still needs normalized dates, negative values for the credit note, and a clean CSV or Excel file before the data is ready for review and import. That is the kind of document messiness that pushes teams beyond basic OCR.

With Invoice Data Extraction, you can upload mixed PDF and image batches, give prompt-based instructions for exactly what to extract, and return structured Excel, CSV, or JSON files that can be reviewed before import. That matters for Polish groups processing English- or German-language supplier invoices, accounting firms handling multiple client document styles, and AP teams that need line-item detail rather than just header values. The platform supports multilingual financial documents, can extract detailed rows, and lets you define rules for formatting, defaults, and document handling, which is closer to invoice data extraction than basic OCR. If you are comparing options in this category, it helps to look beyond OCR alone and evaluate invoice extraction software for ERP invoice imports as a separate workflow layer.

You likely need an extraction-first layer if you need:

• line-level export instead of header-only capture
• custom columns or field rules before import
• pre-import review before posting into Optima
• multilingual normalization across supplier formats
• mixed-batch handling for invoices, credit notes, and attachments

A practical way to think about it is this:

• Use native Comarch OCR first when your remaining non-KSeF invoices are fairly uniform and you mainly want faster entry inside the ERP.
• Use an extraction-first layer first when the bottleneck is standardizing foreign, multilingual, or line-item-heavy invoices before they hit Optima.
• Use both together when you want Optima as the accounting system of record, but you do not want its intake step carrying the full burden of document interpretation and normalization.

If that comparison sounds familiar, the same pattern shows up in other ERP environments too, which you can see in how another ERP handles native OCR versus simpler capture options. For many Optima users after KSeF, the decision is not "Does Comarch support OCR?" It is "Do the invoices still outside KSeF need flexible extraction and standardization before import?"

---

Choose the Workflow That Matches Your Remaining Invoice Mix

Once you narrow the problem to the documents outside KSeF, the decision is less "OCR or no OCR" than where you want the standardization work to happen: inside the native Optima capture step or before import.

Comarch OCR is usually enough when:

• Your remaining documents are mostly standard Polish supplier invoices with familiar layouts.
• You want to keep capture and purchase invoice import as close as possible to the native Optima workflow.
• You only need header-level fields and basic totals, not deep line-item extraction.
• Your team can tolerate some manual correction on exceptions because the volume is modest.

In that setup, the goal is straightforward: scan, validate a few key fields, and import invoices into Comarch ERP Optima without adding another review layer.

An extraction-first workflow is the better fit when:

• You receive foreign-supplier PDFs, multilingual invoices, or mixed-format attachments from shared service centers and overseas vendors.
• You need reliable line items, tax breakdowns, cost-center fields, or custom mappings before the document reaches accounting.
• You want pre-import review, so finance or an accounting firm can check structured data before posting.
• Your downstream process works better when invoice data is exported as Excel, CSV, or JSON first, then mapped into Optima.

That is where Comarch ERP Optima purchase invoice automation often benefits from separating data capture from ERP posting. Instead of asking native OCR to handle every document variation, you standardize the non-KSeF stream first, then load cleaner data into Optima. If you want to compare that handoff model with another accounting stack, see how invoice data handoff works in another European accounting workflow.

After KSeF, most teams do not need to redesign every invoice process. They need a clear rule for the non-KSeF documents that still arrive as unstructured supplier data.