Frequently Asked Questions
About Snowfox and Its Value
Snowfox significantly reduces the manual work involved in coding and routing non-purchase order (non-PO) invoices by automating these tasks using AI.
Customers typically achieve 75–90% automation, leading to faster invoice processing, fewer payment reminders, and better utilisation of early payment discounts. On average, manual processing costs €7–10 per invoice. Snowfox can cut this by up to 70%, saving both time and money.
Additional benefits include:
- Improved accuracy in invoice coding
- Reduced workload for accounts payable and bookkeeping teams
- Fewer supplier queries regarding invoice status
- Greater transparency and control over financial processes
Unlike rule-based systems, which require time-consuming manual setup and frequent updates, Snowfox uses self-learning AI trained on your historical invoice data.
Key differences:
- No need to manually create or maintain rules
- AI adapts to changes automatically
- Significantly higher automation rates (75–90% vs ~25%)
- Improved accuracy through continuous learning and feedback
Getting Started: Trials and Onboarding
The trial is a no-cost way to experience how Snowfox works using your own invoice data. It helps assess the potential automation rate and business value Snowfox can deliver.
Trial process:
- Sign an NDA
- Provide 6 months of historical invoice and coding data
- Snowfox uses this data to train the AI (80%) and test it (20%)
- Receive a detailed report comparing AI predictions with your original coding
- Review results with our team
A secure upload link is provided for data delivery. If preferred, alternatives like SFTP can be arranged. Trial data should reflect your production data format (i.e. Peppol XML and invoice image) to ensure accuracy.
Yes, as long as the data structure matches what would be used in the live environment and includes both invoices and corresponding coding information.
Product Features and Capabilities
Snowfox can enrich and predict all coding dimensions you currently use, such as:
- General Ledger (GL) account
- Cost centre
- VAT code
- Project number
- Reviewer
- Approver
- Workflow steps
All AI models are customer-specific and tailored to your setup.
Yes. Snowfox is system-independent and can be integrated via API or SFTP:
API integration: Unautomated invoices are sent from your AP system to Snowfox for predictions.
SFTP integration: Snowfox receives invoices directly, predicts coding, and returns them to your AP system fully coded.
Currently, Snowfox can help recognise whether an invoice is a PO invoice. While it doesn’t yet support full PO matching, this is on our product roadmap as a future enhancement.
Many PO matching issues stem from process challenges, but Snowfox can support data-related improvements.
Implementation and Onboarding
Implementation can take:
- 3–5 weeks if we’ve worked with your AP system before
- 2–6 months for new systems, depending on complexity
To ensure a smooth implementation, we recommend:
- Appointing a project lead for invoice automation
- Involving AP system admins and accounts payable personnel
- Ensuring internal alignment and understanding of AI’s logic
- Engaging any external consultants (i.e. system integrators) early
Your team will also join testing sessions and be trained on the Snowfox admin tools and analytics dashboards.
Ongoing Support and Customer Success
Ongoing support is included in your annual fee and may include:
- A dedicated Customer Success Manager
- 1 to 12 support meetings per year (based on your needs)
- Access to our customer support portal
- Training and assistance with analytics tools
Ongoing support is included in your annual fee and may include:
- A dedicated Customer Success Manager
- 1 to 12 support meetings per year (based on your needs)
- Access to our customer support portal
- Training and assistance with analytics tools
Security, Compliance & AI Performance
Security is a top priority. Snowfox supports clients in highly regulated industries, including banks, insurers, and public sector organisations.
Protocols include:
- ISO 27001 certification
- Annual penetration testing
- Customer audits upon request
AI performance depends on the quality and consistency of your historical invoice and coding data.
Factors that can impact accuracy include:
- Sparse data from rarely used dimensions
- High variability in suppliers or invoice formats
- Free text fields or rapidly changing values (i.e. project numbers)
Snowfox provides tools and services to manage these challenges, including regular data updates and ongoing model training.
Pricing
- Fixed implementation fee
- Annual maintenance fee (includes support)
- Transaction-based fee per correctly predicted coding dimension
Customers typically see a 50–70% reduction in invoice processing costs through automation.
Contact information
Finland Office
Kaisaniemenkatu 1 B, 6th floor
00100 Helsinki
Sweden Office
Convendum, Västra Järnvägsgatan 3, Floor 9, 111 64 Stockholm, Sweden
United Kingdom Office
London, England
Contact customer support
support@snowfox-ai.atlassian.net