Challenge
Invoicing is the lifeblood for companies and any problems with this function can directly impact cash flow. Valcon’s client, a damage recovery company, which works with insurance companies to repair properties and assets following fires, floods and storms, faced ongoing issues with its invoicing process. Key details were frequently overlooked and it was a big undertaking to manually match cost lines on ledgers and invoices in large volumes. The review process – handled by a large team of financial professionals – was taking a lot of time, leading to lost revenue and delays in payments.
Another problem was the inconsistency of invoicing. Each customer had a unique agreement for waste removal, for example, with one being charged by the hour, another by the bag and a third on a flat rate. This made it challenging to identify missing invoice lines, which had a direct impact on our client’s revenue stream. Our client needed to improve the quality of its invoicing and the efficiency of its accounts team.
Solution
We developed a machine learning system built on a modular neural network architecture to do just that. The AI solution – which uses algorithms that mimic the human brain – identifies missing invoice details and automates validation. The Valcon team focused on this step because it was the most time-consuming part of the process and had the highest potential for improvement through AI automation.
To deal with the variety of invoice processes, Valcon uses ‘autoencoders’, a type of neural network that helps to compress and translate data. In this context, autoencoders act as a ‘translator’ to convert different customer invoice agreements into a common format. This allowed the team to apply a technique known as ‘transfer learning’, where the knowledge gained in solving one problem (invoicing for one customer) is used to help solve similar problems (invoicing for a different customer).
Benefits
Valcon improved the matching, which increased the accuracy of the process, saved serious time and boosted top and bottom line performance. Stand out benefits include:
- Improving accuracy: the AI system reviews the invoices and if it discovers any missing lines, it prompts the user. This increases invoice quality whilst keeping financial experts at the centre of the process, minimising the impact of any potential mistakes made by the AI system.
- Revenue boosted by 3%: as a result of solving issues such as ensuring the right lines are included in invoices and improving all the inaccuracies in invoices, the firm has managed to boost its revenue streams by 3%.
- Higher client satisfaction: the lead time for invoices has dropped from up to three weeks to just a few seconds, which has led to higher client satisfaction amongst the firm’s insurance company clients.
- Achieving 4x monthly ROI: the AI system generates four times its operational costs (e.g. cloud hosting and real-time invoice processing) in revenue each month, demonstrating exceptional financial efficiency and delivering significant value to the client.
If you’re facing similar challenges with your invoicing or other internal processes and want to explore how AI can improve your business processes, get in touch with us at Valcon: [email protected] and [email protected].
