Customer Case Studies

Learn how a range of companies took advantage on machine learning process automation 

SystemsAccountants achieved straight-through-processing in Accounts Payable to enable accounting resources to be re-directed to credit control at a time of crisis.

Other customers achieved productivity gains to re-direct accounting  expertise to development of new financial products, or used new invoice matching methods to enable complex raw materials to be matched to retail products, resulting in higher levels of control over perishable stock, or automatically matched complex projects to enable even finer gain control over expensive R&D investments.

 

Systems Accountants - Straight Through Processing in AP

Straight Through Processing in Accounts Payable is achieved for the first time in any company processing incoming invoices with the result that precious accounting resources are re-directed to important credit control work

Customer Challenge

 

This multi division resourcing and systems integration company wished to improve performance in the accounts payable department. Staff were complaining about tedious manual processes and management were concerned that people would leave. Finding replacements and training them would require management time and cause knock on issues with financial controls. The Finance Director was also concerned that any changes to AP processes could lead to inaccurate financial records.  For example, the finance system was integrated into a third-party timesheet application and changes here could have wide ranging impacts. The challenge was to eliminate manual tasks without changing AP processes.

Solution Provided

Kanbina conducted a requirements analysis and recommended a web-based AI machine learning solution with an API to connect to the finance and timesheet system. The solution would carry out the data processing workflow and tasks automatically. It used real time decisions created by natural language processing and machine learning models. The process was orchestrated by Logic Apps. An API would enable the records from finance systems to train the machine learning models and the decisions to be synchronised in real time.  Power BI was set up to allow management to oversee the performance of the machine learning models.

Impact of Solution

Kanbina successfully delivered the solution and the processing delays in the AP function ended. The machine learning models were retrained every night so that the process continuously improved and 90% of tasks were automated. The staff in the AP function could conduct the processing with speed, accuracy and with minimal delays. Management found that they could use the time saved for more challenging accounting tasks in other areas of finance. Machine learning technologies are now a trusted method and the department now wish to carry out automation in other areas such as Bank Reconciliations.

 

©2020 Kanbina