Connect with us

APPs & Games

Gen Amex ML Based Field

Published

on

Gen Amex ML Based Field is an application that combines the power of rule-based systems and new-school machine learning techniques. It predicts what your customers need. And it was ready to go live months after its initial development.

It was ready to go live months later than originally planned

In its quest to keep customers happy, Amex is a big spender when it comes to both hardware and software. This is where the magic of artificial intelligence comes in. The latest addition to the American Express tech stack is an all-encompassing fraud detection system based on machine learning. Using billions of data points from a plethora of disparate sources, the machine can generate a decision in milliseconds. Combined with a slew of data points spanning over a trillion dollar in transactions, the machine is a formidable adversary to even the most capable of us. If a major player can wrangle this much data in a timely manner, it’s bound to be a winner in the consumer space.

A new app will debut later this year. The company is also looking to rebrand itself as the prepaid credit card of choice. From its early days in the credit card game to its present day state of the art, Amex has a history of maintaining the industry’s lowest fraud rates, which it uses to boost profits and customer satisfaction. The company will continue to leverage its machine learning and data science capabilities as it looks to build its mojo into the foreseeable future.

It’s a blend of old-school rule-based systems and new-school machine learning techniques

American Express is committed to improving its ability to combat fraud, but it also wants to protect customer experience. The company has been open about its work to integrate machine learning into customer-facing functions. It has created an AI lab, called American Express AI Labs, to help advise on AI-based products. This department is responsible for research on natural language processing, cloud platforms, and machine learning.

Machine learning is a computational technique used to generate increasingly accurate relations between feature data and response data. Typically, it is iterative, meaning that the model is constantly refined based on comparisons between predictions and actual observations. Some techniques include support vector machines, artificial neural networks, and decision trees.

Among the most common techniques is clustering. It identifies clusters with similar characteristics, and can also classify data by feature space. Using these techniques, the Gen X model at AmEx evaluates over 8 billion transactions a year.

Another approach is Bayesian network modeling. In this method, variables are modeled as latent variables. This method allows for more intuition on the drivers behind a particular market impact. Other important variables to consider include liquidity, trade sign, and best limit order prices. Identifying these variables can reduce the complexity of the model and reduce transaction costs.

It’s a single model for both international and US markets

One of the best uses for artificial intelligence is identifying and avoiding fraud. The best part is that a credit card company isn’t just going to let a crook get away with the big score. Instead, they are developing science to determine whether or not a transaction is a legitimate exchange of goods and services. Using this scalable model, AmEx has managed to keep their fraudulent rates below industry average. And, in the process, they have managed to keep their customers happy.

To that end, they’ve created a Gen X ML-based field that’s able to take in data from more than $1 trillion in transactions to make a decision in milliseconds. The most interesting aspect of this technology is that it’s a single model that can be deployed for both international and US markets. This makes AmEx the most agile credit card company in the game. Moreover, this model also allows AmEx to take advantage of the latest advances in machine learning and artificial intelligence.

It predicts what customers need

The Gen X model has the power to make decisions in milliseconds. It was built on over $1 trillion in transactions. It has generated over 8 billion decisions for Amex. In addition to making the most informed decisions for customers, the Gen X model helps monitor American Express’ fraud risk. For this purpose, it is able to segment customers based on the level of fraud they pose to the company. This technology also applies to international markets.

The biggest model within Amex’s operations, the Gen X model was created by evaluating eight billion transactions annually. As it ingests data from each transaction, the model considers customer account history. A series of thousands of decision trees is executed, each with its own weights. These weights are then compared against real observations to refine the model’s accuracy.

 

Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *