In today’s fast-paced business environment, efficiency and accuracy are paramount for any accounts payable department. The advent of artificial intelligence offers new possibilities automating repetitive tasks, minimizing errors, and streamlining approval processes. Rillion AI stands at the forefront of this transformation, delivering a powerful solution to automate account coding and generate intelligent workflow proposals—based on your organization’s unique invoice history.
This guide is designed to help you quickly set up and begin working with Rillion AI, ensuring a smooth transition towards a more automated and future-proof way of handling invoices. Whether your goal is to reduce manual intervention, increase transparency, or enhance compliance, Rillion AI empowers your team to focus on higher-value activities while the system handles the details.
First thing first
Does your organization have a license for Rillion AI?
AI is an add on module and can be added to all Packages.
If you are unsure if you have AI included in your Package, please contact Rillion Support or your Customer Success Manager.
The AI product is activated once the agreement is in place.
Model Creation: A model is developed using historical invoice data. For new Rillion customers, without existing historical data, there will be a waiting period - typically two to three months - before sufficient data is available to generate the model.
Inform internally
You are introducing AI capabilities for account coding and workflows to make your AP processes future-ready. Let your team and organization know and ask for their feedback on the new features. Please note that approvers are still responsible for reviewing AI proposals, as errors can occur.
AI candidates
Consider whether to implement a complete rollout or begin with a limited number of companies or vendors.
Start with the most common scenarios
All companies will have different types of invoices, account coding structures and approval hierarchies. Some invoices will follow a standard process while others will be more complex and specific. In our experience focusing on the most common scenarios first will give a great platform to start with.
Pairing AI with a simple reference matching process allows companies to rapidly bring automation to the most common invoice processing scenarios.
If your organization has recently started using Rillion, we recommend first setting up a basic reference matching process. This will generate suggestions for invoices until the AI has gathered sufficient historical data to make its own recommendations. We recommend activating the AI from the start in manual matching mode. Once enough data is available (typically within 2-3 months), you can activate automatic AI matching.
Activation
Under Invoice Log in Company Settings, you can enable AI for each company and select a specific level of automation.
While several settings are available, you should focus on these for activation:
- Enable AI matching for Invoice Posting
- No – AI is not activated
- Only Manually – AI support is enabled for the company, but it must be requested manually for each invoice. The Accounts Payable team can make this request directly from the Invoice Log.
- Automatic – AI is activated and will be done automatically for all invoices that are a candidate for AI. The Coding and flow proposal will be set when the invoices arrive at the Invoice log.
This setting enables AI for the Company.
Our recommendation is to use option C, Automatic.
The Automatic option is designed to maximize automation and functionality of the AI feature.
- Confidence level to set Transfer for AI Matched Invoices
a. All
b. Medium and High
c. High
When the AI determines the coding and workflow, each invoice is assigned a confidence level. This setting specifies the minimum confidence score an invoice must achieve to bypass manual review in the Invoice Log.
We recommend starting with a High confidence level and later allowing invoices with a Medium confidence level to bypass the review.
To enable AI functionality for generating Flows for incoming invoices, select the checkbox:
- Use AI for Flow Predictions
Follow the Step-by-Step guide "AI - Company settings to Activate AI in Rillion Prime" to get started
Start using AI
If AI matching is set to manual for your company, you must specify for each invoice whether to apply AI matching. This action is performed on the invoices in the Invoice Log.
Open the Invoice in the Invoice log.
- Set Type of Matching to AI matching.
- Click Match.
- The Invoice is sent to the AI model for Coding and/or Flow suggestions; this usually takes under 1 minute. You can continue to work with other invoices during this time.
- Reopen the invoice to view any AI suggestions and their confidence levels for Coding and/or Flow.
Follow the step-by-step guide "AI - Manual AI Matching in Rillion Prime" for more information.
Automatic AI matching: When Automatic matching is set on the Company.
- The invoice log displays all invoices processed by AI. Both list and detailed views present the overall confidence level. Additionally, for each dimension and role suggested by the AI, the corresponding confidence level is clearly indicated.
- Confidence levels
Green = High
Yellow = Medium
Red = Low
Changes can be made to both the Coding suggestion and the Flow.
Any changes made, both in the Invoice Log and out in the Workflow, will be reported back to the AI that will keep learning
To change the account coding, follow the step-by-step guide "AI - Working with an Invoice Matched by AI: Adjusting the AI-suggested Account Coding"
And to change the flow proposal, follow the step-by-step guide "AI - Working with an Invoice Matched by AI: Adjusting the AI-suggested Flow"
Technical limitations for AI account posting
AI-powered account coding will handle the 95% most common invoice patterns. In practice, this means if 95% of a customer's historical invoices have fewer than three lines of coding, the AI will provide support for up to three lines. Invoices with more than three lines will not receive a prediction because the AI model cannot reliably work with data representing less than 5% of the total.
Technical limitations for AI Flow proposals
By default, the AI will only suggest roles that have been added into a flow through the following:
OriginOfRole (can be seen by mouse over in the Flow for an historical Invoice):
3 - Role manually added in flow
4 - Role manually added through report
13 - Flow copied from historical invoice
14 -Manually added flow proposal in flow
15 - Manually added flow proposal through reports
16 - Flow proposal manually added in invoice log
18 - Flow proposal from Vendor
19 - Flow proposal from company
20 - Flow proposal reference 1
21 -Flow proposal reference 2
22 - Flow proposal reference match
This means that if the flows you have today are built based on, for example ‘Rule based Flow auxiliary’ or ‘Dynamic flows’, those will not be used by AI to create new flows. AI will also not add the Manager roles into the flow, the manager roles will still be added to the final flow but with the help of normal flow proposal functionality, this means that the Manager role will not be visible in the Invoice log, it will be added at transfer from the log if needed.
AI predictions available for the approvers
When an invoice is sent to approvers, AI-generated information is also accessible for active invoices in their inbox. The organization should be informed that AI functionality has been implemented, along with an explanation of its impact and the related expectations.
If the account coding and workflow are correct, the approver only needs to approve the invoice. If changes are needed, they can be made using the same procedure as before AI was implemented.
The main difference for approvers is that the coding proposal will display a color and percentage to indicate the confidence level of the AI suggestion. If the coding is changed, the approver's values will be applied, and the color indication will be removed.
Level Up: Unlock more values
After using the AI feature in Rillion Prime for some time, it's helpful to review its performance. Identify what works well and where improvements are needed, then adjust the AI setup accordingly.
Certain types of invoices may present challenges for AI systems because of their complexity or variability.
Disable AI for specific Vendors
For some Vendors, the coding should always be the same or might be split over several lines with a fixed percentage. For those Vendors it might be better to use Reference Matching in Rillion Prime instead of using AI.
- Go to the Vendor register and find your Vendor
- Under the section Account Posting you will find the settings for AI
- Uncheck the checkbox ‘Enable AI Matching for Invoice Posting’ to inactivate AI for Account Coding
- Uncheck the checkbox ‘Activate AI matching for the Invoice Flow’ to inactivate AI for building the Flow.
For more information follow the step-by-step guide "AI - How to inactivate AI for a specific Vendor".
Disable AI for specific fields in account coding.
If certain Dimensions are frequently incorrect, they can be disabled from AI prediction. The relevant settings can be found under the Invoice Log section in the Company Settings.
‘Do not use AI predictions on following fields’
Here account posting dimensions that are not to be predicted by AI Matching can be specified. Dimensions are specified using the database title and several values are separated by comma.
For example, if Object3 and Object4 are not to be predicted, value in the field would be ‘Object3, Object4’.
Tips: If you don’t know what Object types your dimensions/object belongs to, go to the Administration – Registers – Objects section, the drop-down under “Object type no” will show the Object types you are using and in the order of Object1-Object8
Utilize AI
To reduce the manual administration of workflows and account posting proposals, best in class companies use AI to complement business logic built up in the system, such as Reference matching.
Improved data on invoice images and captured data makes it easier for the AI to find patterns and make better predictions.
Reference matching to be used when AI generated posting lines exceed:
This setting limits the number of account posting lines that are to be created at AI match. If more than the specified number of lines are created, reference matching will be used instead on the invoice.
Reference matching to be used when AI generated posting lines exceed:
- 1 = use AI if not more than 1 line
- 2 = use Ai if not more than 2 lines
- 3 = use AI if not more than 3 lines
- Always = always use the lines from AI, no limits
Option 3 is recommended. Invoices containing more than three lines frequently receive less accurate suggestions.
Reports and Insights
Analyze the performance of the AP process
In Analytics there is a report called “Invoice Summary” where you have some great KPIs to analyze the performance of the AP process. The report provides details on what matching type is used and how they perform. Use the report to find vendors and invoices that provide poor matching outcomes with the help of AI and investigate potential changes that could be made to improve the matching.
For detailed instructions follow the step- by-step guide "AI - Follow up on AI matched Invoices in Rillion Analytics"
Concluding Remarks
Now you know how to use AI to streamline your account coding and workflow process. This will help you to save time, increase accuracy, and continually improve your organization's AP operations.
Wishing you the best of luck as you begin your journey with the new AI solution. Success is just a few clicks away!