We have progressed from the development stage of artificial intelligence (AI) and machine learning (ML) technology into widespread implementation across industries. This advanced technology is everywhere, but most people don’t know they interact with it daily. From virtual personal assistants like Siri and Alexa to the recommended shows that Netflix serves up based on viewing habits, ML powers these tools we’ve become so accustomed to using — even if we don’t stop to think about how it happens.
What’s the difference between AI and ML?
AI is computer science that is focused on the capabilities of machines to imitate intelligent human behavior. AI allows computers to process large amounts of information and data and provide a computer-generated conclusion. ML enables computers to learn — and also learn how to learn. These advances are revolutionizing modern life and business.
So how are AI/ML tools being used in their next generation of implementation, and what’s next?
Credit Card Fraud
The next generation of AI/ML can combat credit card fraud before you even check out at the register. One recent report showed that consumers lost $6.4 billion in credit card fraud transactions in 2018.
Common types of credit card fraud include stolen credit cards and issues with the cardholder not being present. AI/ML-based transaction screening tools can enable businesses that accept credit cards to identify a fraudulent transaction before it’s completed, which can reduce the number of fraudulent card users attempting transactions. It also decreases the time and resources required to recoup revenue from the financial institutions that handle those transactions.
As reported in MIT News, the big challenge we face with using machine learning to detect fraud is dealing with false positives. To help reduce the challenge of false positives, researchers at MIT’s Laboratory for Information and Decisions Systems (LIDS) developed an approach that reduces false positives by 54%.
MasterCard has developed an in-house solution called Decision Intelligence, which is a real-time authorization decisioning solution that applies thousands of data points and modeling techniques.
AI/ML algorithms identify specific characteristics that make a transaction more or less likely to be fraudulent. If you have ever been traveling and received a message from your credit card company that there was a concern, it was probably because their ML technology didn’t recognize transactions taking place in a new location with different merchants that were uncharacteristic to your profile.
A Virtual Stylist
What if you walked into a store and were greeted with recommended pieces ready for you to try on based on what you were wearing? With AI/ML technology, this has become a reality.
Some retailers are already piloting AI/ML-based tools that recognize customers’ faces and clothing to make recommendations. In Hong Kong, fashion retailer Guess opened a pilot FashionAI concept shop at Hong Kong Polytechnic University. At the concept shop, machine learning and computer vision are deployed to “learn” from consumers and designers within the system. Customers checked into the concept store with facial recognition technology. RFID-enabled clothing rack options automatically showed up on the smart mirror, which offered styling suggestions.
Other AI/ML-based styling assistants provide the information to sales associates so they can personally provide customers with recommendations, making the shopping process more seamless and efficient.
SMART Master Data Management
Data management and duplicate data entries have always been a struggle for businesses of all sizes.
Think about making an online purchase, such as ordering pizza for takeout. You log on to the website, enter your information to create an account, complete your purchase and move on with your dinner. The next time you visit that site for pizza, you’ve forgotten your new login information. You simply create a new account.
From a user’s perspective, that’s no big deal. But for business owners, that type of multiple data entries can become a nightmare.
A database with 30 million users might actually only be three million unique users. This is an issue that has caused database administrators countless headaches, and it hurts the bottom line for businesses.
Leveraging AI/ML technology, companies can implement an all-in-one server add-on that runs seamlessly in the background, scanning and analyzing user entries in real time. It can even be configured to block duplicate user sign-ins as they happen. The AI/ML solution does this by automatically matching user data and comparing data points such as username, email, phone number, address, Social Security numbers, linked credit cards, IP data and more. There is no need to run custom queries or reports, saving time and human capital.
Equifax was challenged with inconsistent customer data that was housed across 20 different systems, including billing and fulfillment, which comprised 2.1 million customer records. It integrated data from 45 million businesses and created one ID number for each customer. Equifax gained a more clear picture of the customers’ full relationship with it, including how much revenue each product and each customer represented, enabling better decision making and cross-selling opportunities.
These solutions can be installed on any server, without the need for expensive and costly custom development.
These are just a few examples of the way the world continues to embrace AI and ML technology to innovate the way we live. The organizations that embrace these next-generation applications will operate more efficiently, provide their customers with better experiences and lead their industries. Business leaders should make sure their organizations don’t get left behind.
All Rights Reserved for Forbes Editors’ Picks