How can AI Inside be used?
AI Inside offers two main ways in which it can be used:
- Background Automation and
- Progressive Automation.
1. Background Automation
Involves using AI Inside as a background process that provides accurate and timely predictions for your accounting system. This feature allows AI Inside to analyze accounting data and suggest the most likely entries for the purchase invoices, making the invoice handling process more efficient and streamlined. The pre-filled posting suggestions provided by AI Inside can greatly reduce the amount of time and effort needed for manual data entry, freeing up the accounting team to focus on higher value-added tasks.
2. Progressive Automation
Takes things to the next level, as it allows your company-specific AI models to bypass the accountant in the whole invoice handling process. With this feature, AI Inside can accurately estimate both prediction accuracy and the invoice relevance, ensuring that the accountant is shown only the invoices that need to be seen. Up to 90% of the invoices can be processed automatically all the way to the accounting ledger, saving the accounting team even more time and reducing the risk of errors.
Possible next steps
- First, determine which AI Inside features will be integrated with your software. This may include pre-filled posting suggestions (=Background Automation) or bypassing the accountant (=Progressive Automation), as well as any custom features that have been developed by the AI Inside team.
- Next, consult the AI Inside Technical REST API documentation provided by FabricAI to determine the API endpoints and authentication methods that will be used for integration.
- Develop a script or software module that can communicate with the AI Inside REST API using the appropriate API endpoints and authentication methods.
- In the script or software module, define the appropriate data structures and variables needed to receive and send data to and from the AI Inside REST API.
- Test the integration by sending sample data to AI Inside and ensuring that the correct predictions or suggestions are returned.
- Once the integration is tested and validated, deploy it in the production environment.
This approach can provide a streamlined and efficient way to integrate AI Inside into accounting software, allowing for powerful automation capabilities and accurate predictions.
In order for AI Inside to be effective, it requires structural invoice data to train its deep learning AI models and produce accurate predictions. Currently, only a few markets have a high adoption rate of electronic invoice formats, and many still rely on PDF invoices (which are NOT structural e-invoices per se).
AI Inside processes data in .JSON format, and if you need help connecting the data that you have gathered from an OCR provider that scans PDF invoices into the correct AI Inside format, please contact us. We are here to assist you and ensure that you can leverage the full power of AI Inside to optimize your accounting processes.
AI Inside currently supports the following invoice formats
- TEAPPSxml 3.0
- Finvoicexml 3.0
- EHF 3.0
- PEPPOL BIS 3.0
- VISMA Smartscan OCR data type