Advanced machine learning helps improve current processes in cash management and goes beyond them for new business value. What does it take to harness the advantages of Smart AI?
We are in the midst of a fast-paced, complex data-driven world, fueled by machine learning. The next industrial revolution is that of smart-software which will provide its recommendation using artificial intelligence (AI) technology facilitating our decision making.
Up until now we have focused on surviving instead of improving. Companies have extinguished the fire that burns the most at the moment. But few strategic decisions by a C-level executive today could do more harm for a company than not investing in new technologies and solutions. The key to staying competitive, and what should also be the cornerstone of every organizations´ business model, is full centralization, together with enhanced user experience, and using a single platform that provides full visibility of the entity's cash flows and payment streams, in almost real time.
It enables companies to start adopting solutions with a high degree of advanced machine learning called Smart AI – which we consider to be the ultimate form of AI.
Focus on adding real business value
Centralization is basically about standardizing data. You harmonize the processes and systems by bringing all transactions to the same time zone, and into similar data structures and formats. What makes a big difference today, as a result of the rapid pace of the digital development, is the amount of data we now have the power and possibility to analyze.
Using new Smart AI-driven technology, machine learning can analyze historical data points and make it possible to refine and increase effectiveness of existing process. Subsequently, it even allows us to identify business values we did not know existed in our company.
For example, manual procedures such as adding posting dimensions to supplier invoices could be done more efficiently and more accurately with machine learning. Automated posting further allows to uncover inefficiencies in the organizations’ chart of accounts; a new business value which was not easily identifiable in the manual posting process. Similar analogies can be made for account receivable processes.
Avoid becoming a victim to your own overhead costs or fraudulent behavior
Many of these basic cash management and extended purchase-to-pay processes are still handled manually, which lacks the ability of scalability. Today, key challenges for any CFO or Group Treasurer are the lack of visibility of who has payment rights on bank account level, the usage of multiple and outdated IT-systems, a plethora of bank accounts in too many banks around the world, and financial policies that are misinterpreted or ignored.
On a global scale with many thousand bank accounts, getting a full, updated view of available cash can be a real challenge. It is also far from realistic to demand daily manual reports from subsidiaries.
One in four companies in the Nordics are subject to cybercrime (PwC Global Economic Crime Survey 2016). That type of criminal behavior is the second most common financial crime, after classic fraudulent behavior and embezzlement.
It is proven that machines excel at crunching and finding recurring patterns in massive data collection. In the area of payment fraud, advanced machine learning is capable of doing the reverse. It can spot unusual patterns in historical payments. These patterns can be complex or even simple, but in any case invisible to the human eye. A smart software will alert the human about potential anomaly before the payments are released to the bank, and before the damage is done.
Even though the digital maturity among companies varies, we should aim to strengthen digital security and compliance. This is made possible by having controlled access in place, logging and storing digital fingerprints and historical data, and finally enabling Smart AI to detect unusual behavior based on learning from historical data.
The very first step in this journey is to centralize cash management and create a single system that enables you to handle all the financial processes in the same ecosystem.
An informed workforce stands prepared for Smart AI
Organizational aspects of digital transformation are often forgotten. We need to explain the advantages of new technologies and anchor the idea of it among our colleagues. By centralizing our financial processes, we are taking a giant first step toward augmented workforce, where man and machine work side by side, adding real value to our business.
Competence in new technology, and knowing how to work in an agile way and how to interact and benefit from the recommendations offered by the intelligent machines will be the key differentiators.
Smart AI, driven by advanced machine learning does not mean much if we do not manage to create a well-informed culture among our colleagues at the same time. By making them aware of these new technologies, familiarizing them with the idea of Smart AI, and putting effort in educating them, companies can retain, improve and evolve the most fundamental financial processes in a way that is most fulfilling for the entire workforce. As a result, they are able to harness full advantages of the dramatic progress in Smart AI, and the real business opportunities it offers in today’s digital economy.
How can Machine Learning and centralization help Treasuries and SSCs in their daily Cash Management work? Watch 20 minutes introduction to learn more.
Karl-Henrik is a passionate Cash Management professional with background as a Cash Management Advisor at a large Swedish bank followed by six years as a Cash Management Director at a Treasury department in a global multinational. Educated in Finance but with a "techie" mindset he is often seen speaking to his smartwatch or discussing disruptive Fintech with like-minded.
Ali has extensive experience in data science and creating advanced Artificial Intelligence solutions in various forms. At OpusCapita he is responsible for designing, creating, and evaluating disruptive data-driven intelligent systems. He serves as a reviewer for numerous international journals in applied machine learning and maintains collaboration with research institutes.