By Emma Hitzke, Senior Director, Emerging Technologies, Oracle
If you’re confused about how to adopt emerging technologies easily and cost-effectively, you’re not alone. With artificial intelligence (AI) and machine learning (ML) on everyone’s minds these days, the constant threat of disruption from the next big thing can be paralyzing.
But, like any technology, AI and ML are tools that can solve a business problem. By starting with your business needs instead of focusing on the technology, you put the business value at the center of your transformation efforts. For example, if you are looking for ways to automate repetitive processes such as cash disbursement and revenue management, or to close the books more quickly, efficiently, and accurately, then an ERP solution with embedded ML can deliver on these promises. In fact, research from the McKinsey Global Institute concludes that 40 percent of finance activities can be fully automated, and another 17 percent can be mostly automated.
The bottom line is AI can improve automation, predictions, and decision-making while lowering costs and improving efficiency.
The opportunities of AI for finance
When coupled with analytics, AI can provide unprecedented capabilities to better manage the business, make decisions, and discover patterns. AI can improve the finance function in three primary ways:
- Improving productivity, efficiency, and engagement: Next-gen experiences such as conversational interfaces and intelligent user experiences can simplify multiple tasks.
- Automating tasks: Automation not only reduces costs and improves accuracy, but it also liberates people from performing mundane work and allows them to focus on more strategic projects.
- Improve predictions: ML (a form of AI) can increase the accuracy of forecasting and budgeting which, in turn, supports more informed decision-making.
The challenges of AI for finance
Advanced technologies cannot bring about transformation on their own, however. They require a receptive culture and the ability to provide highly relevant data for analysis. Among the challenges to a successful implementation of AI are:
- Improving communication: If your business and IT teams aren’t communicating already, adding AI won’t improve the situation. The process begins with level-setting one another’s expectations: Share business objectives and clarify the scope and timeline of any new AI initiative. You must also consider the cultural shift that new technology creates.
- Finding the right data: Even the best algorithm cannot provide meaningful insight if the data you start with isn’t of the highest quality and relevance. You need to have data-cleansing processes such as de-duplication, and plan to augment data you’ve collected with additional information derived from internal and external sources (for example, combining internal financial data with internal HR data or external business bulletins). Most importantly, you must liberate your data from functional and operational silos across the organization.
- Assuring security: The explosion of data held in corporate systems is already raising red flags around security and privacy concerns. Organizations must learn how to maintain the security (not to mention anonymity) of the data they collect and process in order to retain customer trust and comply with regulations.
- Acquiring new skills: The drive to mine data for insight has made data scientists a hot commodity. Once your finance team acquires the skills required to extract meaning from your data, it will be able to derive the full benefit of AI.
The solution: AI out of the box
When software solutions come with AI already built in, your organization can embrace the latest innovations right away to improve user engagement, collaboration, and performance. At Blockland 2018, Oracle CEO Mark Hurd told the audience, “I do not believe AI will be a, quote-unquote ‘application.’ I believe it will be a capability integrated into all applications. Chatbots, digital assistants—there won't be a chatbot application. There will be chatbots integrated into all of these applications.”
Oracle Cloud applications (such as Oracle ERP Cloud) include ML capabilities embedded within familiar user interfaces and business workflows. They enable customers to realize quick value from the latest innovations in AI, conversational interfaces, natural language processing, blockchain, and the Internet of Things (IoT). Consider the ways in which AI can make these three finance processes more efficient:
- Order to Cash: AI can automate receipt allocations by intelligently matching receipts to payments. It can also assign and adjust customer credit ratings, allowing your business to assign different payment terms and strategies to each customer, and it can help establish customer-specific collection priorities and strategies, such as reserving aggressive collection practices for less strategic customers.
- Procure to Pay: AI can also automate and optimize many processes within the accounts payable function. Automatic invoice matching and allocation can reduce invoice holds by matching invoices to POs and properly allocating them to the chart of accounts. AI can also inform early payment discounting or dynamic discounting programs by setting optimal rates for individual suppliers. It can also minimize supplier risk by drawing from third-party data (such as credit ratings) to help your company predict anomalies and negative conditions with suppliers before they become an issue for your inventory and production processes.
- Accounting and Financing: Smart automation can streamline entry and classification of expense information as well as audit expense entries for anomalies to save on manual audit efforts and reduce error and fraud.
AI offers companies tremendous opportunities in the near future. Right now you can leverage continuous innovation and advanced technology already built into best-in-class SaaS solutions from Oracle Cloud. Learn more at Oracle.com/ERP.