Working Capital

AI is Ushering in a New Era of Integrated Receivables

Sayid Shabeer of HighRadius explains how AI engines can tackle manual processes around cash applications, credit management, and more.
Vincent RyanOctober 1, 2019

The idea of implementing AI as part of a business process is often received with mixed feelings. Many believe that AI will take over jobs — or even entire industries. But in reality, fears may be overblown. Jobs, as we know them, may change, but adapting to innovative technologies will free up time and resources to focus on critical tasks. We saw this happen before when mobile disrupted the landline, but it opened up opportunities for mobile business, innovative app development, revolutions in gaming and so on.

When it comes to integrated receivables today, AI engines can tackle vast amounts of manual processes that overwhelm employees, including processes around cash applications, credit management, collections management, and dispute management. And there’s still plenty of room for new ecosystems of innovation, including chatbots and digital assistants, that can offer customized customer management and AI systems to reduce fraud.

Drive Business Strategy and Growth

Drive Business Strategy and Growth

Learn how NetSuite Financial Management allows you to quickly and easily model what-if scenarios and generate reports.

In a new era of integrated receivables, here’s how AI can impact four specific activities in cash application, credit, collections, and deduction dispute management.

Remittance Data Capture

When companies process e-payments, there’s often a decoupling between payment and remittance, which is a painful nuisance for accounts receivables teams. Even with robotic process automation (RPA) in place, remittance data capture requires many labor-intensive hours and working with an unwieldy database of information. Personnel are frequently tasked with manually rifling through a variety of sources to compile remittance information.

Luckily, an AI system that employs optical character recognition (OCR), natural language processing (NLP), and machine learning can enhance the human-machine relationship. It can capture raw remittance data found in email bodies and attachments, EDI files, check stub images, and buyer web portals, while classifying and segregating out non-remittance data. It can then match payments data with the captured remittance data and open invoices to straight-through process and closeout open accounts receivables.  AI can also help an analyst handle exceptions quickly and efficiently with machine learning-based suggestions to resolve exceptions.

Payment Date Predictions

Typical collections processes are reactive. Activity is often only triggered for a collector once an invoice due date passes, and customers are only contacted once the payment is already late. This ultimately results in slower cash conversion cycles, as well as facilitating inefficient processes that run up operational costs.

An AI engine can intelligently create payment date predictions, which can help receivables teams prepare for collection activities or fine-tune their collection strategy before an invoice is due. It does this by implementing machine learning algorithms that process factors such as payment trends for an individual customer, number of payments delayed, number of promise-to-pays broken, and total open invoice amount.

With that prediction in place, analysts can spend more time focusing on accounts that have and are expected to have large delayed payments. In other words, collectors do not have to wait for an invoice to be past due to take action, but rather can work off of predicted delays, and offer discounted payment for early payments.

Dispute Validity Predictor

Every year, companies lose large amounts of money in the form of mismanaged or unresolved deduction disputes. It’s up to your team to identify the small percentage of invalid deductions among all the deduction claims that come through. This is why deductions research can aptly be described as looking for a needle in a haystack.

An AI system can predict the validity of these disputes and deductions, saving massive amounts of time. AI can use factors like dispute history, cash discounts versus invoice discounts, product categories and value percentages, proof of delivery (POD) records, among others which can then identify patterns and predict the validity of future deduction claims. Years of deduction resolution data is leveraged to train machine learning algorithms. Then, instead of spending time searching for invalid deductions, valid deductions can be pipelined through for quick reviews or auto-resolution, while teams can spent their time effectively and deal with potentially invalid deductions head-on.

It’s Time To Embrace AI

The bottom line is that AI is here to help. It offers financial professionals the opportunity to redefine and move up their value and contribution to their company, moving from manually intensive or repetitive work to more critical and value-creating work. It helps get information delivered to the right people, at the right time and empowers better relationships with partners and customers.

In the integrated receivables sector, AI can be put to good use to alleviate huge amounts of stress over remittance management and credit-blocked orders. It can also manage delayed payments and handle deduction claims. Done right, it will foster business agility and fuel growth.

Sayid Shabeer is the chief product officer of HighRadius.

This story originally appeared on

© 2019 Benzinga does not provide investment advice. All rights reserved.