The future of artificial intelligence in finance

FORECASTING has been around since the beginning of human expression. Efforts to predict what will happen in the future have ranged from the writings of Nostradamus to psychics on television offering their services for a fee. And while most of us do not have a crystal ball or the skill of being clairvoyant, we have the capability to predict certain events based on information that is either seasonal or cyclical in nature. When it comes to finances, there are modern methodologies to help us manage the growing amounts of data inputs and outputs.

Accounting software such as QuickBooks helps provide an overview of finances, whether for personal or professional use. In the commercial world, many businesses are now relying on technology driven by artificial Intelligence (AI) to make predictions about future consumer behaviours based on algorithms and a robust data set that outlines certain patterns to assist in the budgeting and planning process. But AI can also be used in a myriad other ways.

In South Africa, First National Bank is leveraging the power of AI to flag nefarious activity to mitigate the risk of money laundering and fraud. The system was designed based on a set of in-house rules, models and algorithms that saves up to 70% of the time it would take a human analyst to identify these issues. But the tool itself does not replace the work of human hands. It merely augments their ability to get the job done in a more timely manner. What might have taken days can now take as little as eight seconds to complete.

By creating a more agile system, the bank has brought in more stability, which builds trusts with its customer base. Its efforts to combat fraud not only raise its public profile as a trusted enterprise, but also have a direct, positive impact on personal investors seeking to plan for retirement.

According to a joint report by Microsoft and Ernest & Young, nearly half of South African businesses are already testing out AI within their enterprises. At present, the most common forms are chatbots, robotic process automation and advanced analytics. Over the past decade, the total investment in processes that include planning, scheduling and optimisation has amounted to US$1.6 billion.

Over four-fifths (83%) of those surveyed expect to use AI for automation; another 70% will apply it toward predictive processes. With a more comprehensive overview, businesses are able to better serve their customers.

While many claim data is the new supercurrency, not all data is created equally. Data quality plays a significant role in leveraging the power of AI-driven technology. Three key benefits should be considered when applying AI solutions in finance:

1. Streamlining data preparation. Automating processes can

replace error-prone manual ones, which have an enormous impact on data quality. “Dirty data” – inaccurate or irrelevant data – can lead to large delays in proper data analysis, which can slow down many business functions. AI offers accurate predictions based on consistent, high-quality source data. By streamlining the data preparation process, businesses have the ability to remove irrelevant information through AI and provide more accurate services to the customer.

2. Predictive forecasting.

Given the world’s current complexity and rapid change, businesses require fast access to various what-if scenarios in order to consider best-case and worstcase scenarios. Automated forecasting with advanced predictive technology that is integrated into existing systems heightens accuracy by taking into account certain aspects such as product seasonality and cyclicality. This capacity also extends to the seasonal labour market to determine, for instance, how many staff are required to serve the tourism industry during periods of heightened demand.

3. Machine learning. The beauty of AI-based solutions is their ability to provide more accurate forecasting results using multiple data inputs. Instead of manually assessing individual spreadsheets, easy-to-use AIassisted planning solutions can leverage advanced machine learning capabilities to accelerate the process of turning data into insights.

In addition, AI-assisted planning tools offers solutions that help enhance employee productivity, reduce customer churn, improve supply and demand alignment, and – within the manufacturing sector – actively manage machinery downtime. These AI models provide improved data classification, clustering, feature selection, and more.

Relying solely on heaps of Excel spreadsheets to keep up with today’s consumer demands is no longer a viable way to manage data. In order to improve overall customer experience and remain competitive, organisations must pivot toward new technological solutions that can help decision makers remain agile in the face of an ever-changing business landscape.

The technology exists. It now depends on organisations’ willingness to embrace the inevitable evolution towards technology-assisted efficiency and effectiveness to better serve consumers long term.

Maličević