Automate Invoice Coding in IFS With AI That Learns from Your Data

Snowfox AI learns exclusively from your own invoice data — it is not a shared or global model. The system continuously refines predictions based on historical and incoming invoices specific to each organisation.

Typically, 6 to 12 months of historical invoice data is recommended, along with corresponding GL coding and workflow data (including reviewer and approver assignments).

Cost centre modifications can be communicated through the admin tool for immediate adaptation. If the change is not entered manually, Snowfox will typically learn the new cost centre automatically after processing around 3 to 5 invoices.

Yes. Every invoice — including those that have been manually corrected or rebooked — is used as retraining material. This process reduces future errors and increases confidence over time.

Yes, each Snowfox AI model is trained from scratch for the individual customer — no data is shared between customers. Historical data extraction can be performed by internal teams or IFS service partners using provided documentation.

Customers typically achieve automation rates of over 85% — and often over 90% — within the first six months. Auto-posting is configurable by confidence thresholds per dimension.

Customisation is rarely necessary. The AI learns all coding nuances directly from your historical invoice data, including supplier-specific patterns and exceptions.

Snowfox AI models are always trained per customer — never shared across tenants. Individual customers can maintain multiple unique models if business requirements demand it.

Snowfox is a separate ISV (Independent Software Vendor) add-on and is not included out of the box with IFS Cloud. It functions as an intelligent automation layer between invoice sources and IFS.

All existing automation rules and templates remain configured within IFS. Snowfox works alongside these, adding AI-driven coding predictions on top of the rules already in place.

Each system serves distinct functions. Pagero acts as the delivery layer, receiving invoice images and Peppol files. Snowfox then enriches those files with GL account coding, cost centres, and other dimensions, and determines the correct authorisation routing.

Snowfox is an add-on to IFS, not a replacement for Pagero. Pagero handles the legal delivery of invoices (via image or Peppol), while Snowfox applies the accounting logic.

Snowfox supports all invoice sources. Whether invoices arrive via Peppol or as scanned images processed through an OCR solution, Snowfox can handle the full range of input formats.

Invoices continue flowing through Pagero normally. Snowfox sits between Pagero and IFS, automating the coding and routing process, but does not replace Pagero's role in invoice receipt and delivery.

Yes. By the time an invoice reaches IFS, it has already been processed by Pagero for data capture and enriched by Snowfox with GL coding and authorisation routing.

Yes. Snowfox sits between Pagero and IFS in the invoice flow, enriching the FINVOICE XML with GL coding, cost centres, and authorisation data before the file is passed to IFS.

No. Snowfox is implemented as an external layer and does not require any modification to the core IFS code. This ensures full compatibility with system updates.

Yes. Snowfox integrates as a middleware layer between Pagero and IFS. It receives e-invoice XML files from Pagero, enriches them with coding and authorisation data, and passes the completed invoice to IFS.

Yes. End users continue to work entirely within IFS — all approvals, corrections, and final postings take place in the familiar IFS environment.

Yes. Pagero sends invoice files to Snowfox first. Snowfox enriches each file with GL coding predictions and other relevant data, then forwards the completed invoice to IFS.

There are three main connection points: (1) inbound invoices received by Snowfox from your invoice source (e.g. Pagero); (2) the enriched invoice file delivered from Snowfox to IFS; and (3) completed coding and authorisation flow data returned from IFS to Snowfox for model retraining.

No. Snowfox operates independently and does not require IFS AI to be enabled.

If Pagero rejects an invoice, Snowfox will not receive it, as Snowfox sits downstream of Pagero in the process.

For PO-based invoices that include additional charges — such as freight or packaging — that are not reflected on the purchase order, IFS Cloud 24R2 introduced functionality to automate the handling of these charges.

Coding explainability is a feature coming very soon. Once released, users will be able to see exactly why the AI predicted a specific GL code or cost centre.

Yes. Snowfox can automatically consolidate invoice lines that share the same accounting and tax code, simplifying the invoice structure before it is transferred into IFS.

Snowfox can either complement IFS precoding or independently predict work order and project codes. The system performs well, especially with consistent attributes like supplier addresses.

Snowfox is launching an invoice validation service in Q2, which will address a range of error detection scenarios — including incorrect VAT, missing entity data, and other common invoice discrepancies.

Goods receipt registration for PO-based invoices is handled within IFS as standard. Snowfox is focused on non-PO invoices and does not manage the arrival registration process.

Snowfox does not currently support the preposting of PO invoices as configured within IFS. This is, however, an area being explored for future development.

Yes. IFS Cloud 24R2 includes functionality to automatically detect and add coding for additional charges — such as freight and packaging.

Snowfox is primarily focused on non-PO invoices. For PO-based invoices, IFS handles the matching and coding process natively.

Snowfox is priced on a fixed monthly fee, calculated based on your annual non-PO invoice volume. There are no per-transaction charges.

Snowfox is recommended for organisations processing at least 1,000 non-PO invoices per month. Below this volume, the return on investment may be limited.

Snowfox pricing is based on a fixed monthly fee calculated from your annual non-PO invoice volume — not a per-transaction model.

Snowfox operates on a fixed monthly fee based on your total annual invoice volume. There are no variable per-transaction costs.

The number of legal entities in your group does not affect Snowfox pricing. Pricing is based solely on the total volume of non-PO invoices received annually across your organisation.

The technical integration typically requires only a few days of working effort. In terms of calendar time, the full implementation — including testing and go-live — usually takes between 3 and 6 weeks.

Yes. Snowfox is fully compatible with on-premise IFS installations and functions in the same way as it does in cloud environments.

The typical implementation takes between 3 and 6 weeks of calendar time, with a modest amount of working effort required.

Migration from IFS 10 to IFS Cloud is straightforward with Snowfox. The recommended approach is to implement Snowfox on your IFS 10 environment first, so the AI model is already trained and ready when you go live on IFS Cloud.

Yes, and Snowfox does not charge additional fees for the migration. Moving from an on-premise IFS 10 environment to IFS Cloud is a straightforward process.

No retraining is required. Snowfox provides a comprehensive admin tool where approver and reviewer assignments can be updated directly.

Cost centre changes can be made directly in the Snowfox admin tool, and the AI adapts to those changes immediately.

All configuration data is mirrored directly from IFS, so there is no need to maintain a separate dataset in Snowfox. Ongoing maintenance is minimal.

The supplier invoice data capture solution built into IFS Cloud 25R2 is a potential replacement for Readsoft and Kofax.

The tools share similarities but differ substantially. Unlike Propell, Snowfox does not replace the IFS user interface — it works entirely in the background within the IFS environment.