How to put the old limits on productivity in the past
Invoice receipt in SAP - today this essentially runs automatically in large companies. Usually smoothly and quickly, as long as the relevant information for automatic account assignment is within the invoice data received. Everyone knows that in practice there are examples every day where this is not the case. Then ambiguities arise, the automatic posting fails or - even worse - generates completely wrong allocations.
The barrier to automation is often that the correct G/L account assignment is unclear. In such cases, only slow and expensive manual work has been an option until now. But the automation possibilities of 2021 are no longer limited to the automatic reading of tables. Today, you can push this productivity frontier by using Artificial Intelligence. avvaneo's powerful AI models are already being used in more and more companies, reducing processing errors while providing a significant increase in productivity that would not be possible without AI.
What exactly is happening?
To eliminate ambiguities can be a challenging task. Usually, people are needed because they can understand context and therefore understand what is incorrect in a classical if-then process.
Paradoxically, in the case of incoming invoices, humans are now at a disadvantage compared to a machine solution. The reason: The machine remembers everything and can therefore recognize correlations in all existing data at lightning speed. Above all, it can recognize patterns in thousands of columns and millions of rows of data and draw appropriate conclusions from them - faster than it takes a human being to even look at the entire invoice. While the clerk will start researching the reason and possible account allocation after viewing the invoice, the machine will already have a valid prediction ready.
This is exactly how the avvaneo AI solution is used for invoice receipt. If at any point in the process an invoice is unclear, the AI finds the most probable assignments, practically in the blink of an eye, and either automatically inserts them immediately or suggests them to the clerk as options to pick from. This is especially true for the two primary issues that are important in posting and can easily go wrong:
- G/L account assignment: project reference, customer reference, stocking, etc.
- Approval: Automatic assignment of the invoice to the next recipient
Who should add AI enhancements to invoice receipt processing?
We don't want to promise a fixed percentage to you, but we can say: The effects are measurable and are confirmed time and again in practice. The key parameters to be aware of are the throughput times and number of errors, which has also been confirmed in this project.
In actual practice, the improvement always depends on the initial situation. Generally speaking: The more time you lose today due to manual account assignment and laborious G/L account research, and the more errors slip through due to time pressure, the more urgent it is to recommend: Review the possibilities of integrating our machine learning models into your existing automation.
Aren't AI projects extremely complex and costly?
We can answer this with a very clear "No, not with us". Even for large companies with multi-layered supplier relationships, large departments, many authorized approvers, and millions of invoices per year, we can usually get an AI project up and running in about 3 months.
For a company taking on this task on their own, this is definitely a different picture. Available resources who are well versed in both SAP automation and AI are not easy to find. At avvaneo, on the other hand, we have the perfect qualifications: not only do we know financial automation in SAP inside out, but we also have proven AI approaches in use at our customers.
Because the best AI solutions emerge when human intelligence and experience have first taught the machines what to do.
avvaneo AI overcomes previous limits of productivity
- Companies increase their degree of automation, the administration becomes "fitter" and reduces its costs.
- Root causes for productivity losses disappear completely or can be processed much faster.
- Reduction of misallocations increases cost transparency.
- Budget overruns on individual projects may be more noticeable.
- Third-party costs can be charged on correctly instead of possibly burdening your own company.
- The training time for new team members is reduced because the AI replaces the missing experience, as it were.
