Artificial Intelligence

AI in Theory

Artificial Intelligence (AI) is an area of Computer Science that seeks to simulate human intelligence in computers. Basic activities that people perform every day like carrying on a conversation, finding the way to the grocery store, recognizing a familiar face, or learning from mistakes are all very difficult things to get machines to do. Despite the challenges, researchers in AI have worked to endow computer programs, and by extension the machines that they control, with some, or all, of these abilities. Their efforts have led to the development of ever more exciting and useful technologies that improve our daily lives.

Today, chess programs are capable of beating the best human players, and voice recognition technology allows you to talk your way through a company’s telephone system without needing to repeat yourself. At the same time, self-directed vacuums are able to navigate their way around your home tidying up, and robotic soccer teams play in competitions against one another. In the financial realm, credit card companies and other financial institutions use AI techniques and technology in detecting fraud. While for a wider audience, search engines employ it to improve search accuracy and relevance, and spam filters to detect unwanted email.

Back to top

AI in Practice

Most online budgeting tools available today are limited in their ability to categorize your banking and credit card transactions. They either require you to manually categorize your transactions or they fail to accurately categorize many of them. The difficulty lies in the sparse data found in the OFX specification - the standard for banking and credit card transaction data.

Developing rules to handle the many and various data problems that may arise is time consuming, and fragile. In addition, user-defined rules are typically not good at handling ambiguity, such as when a transaction for a particular company is assigned to different categories. For these reasons, many people either give up on money management tools altogether, or simply do not bother to create more than the most basic rules. Instead they need to manually and repeatedly assign new transactions to the appropriate category.

It should be easier!

With Neuralus, it is. Neuralus has created Neuralus INSITE™ – a dynamic, intelligent categorization tool that learns from user interaction. Here’s what it does:

  • Handles imperfect, ambiguous, and new transaction data, and learns to accurately categorize similar transactions into dynamically-defined categories, and adapt to the addition of new ones. This minimizes the work required of the consumer to keep transactions categorized correctly, and they don’t ever need to define a single rule.
  • Learns from user's choices and adapts to newly encountered transactions automatically.
  • Supports dynamically created categories, which allows users to create as many categories as apply to them in particular without being constrained to a common set of global categories that must meet the needs of everyone.

Neuralus INSITE™ is a smart solution to your customers’ budgeting and expense-side planning needs.

Back to top