How Predictive Analytics Saved me $1,000 at Lululemon

The text arrived at 12:19 PM yesterday. I was in the office commons filling up a glass of water.

Amex Fraud Alert: Did you just attempt a 1,000.00 USD charge on Card ending 11006 at LULULEMON GIFTCARD? Reply 1 if yes, 2 to call Amex.
I pressed 2.

American Express verified my identity and then assured me that I had nothing to worry about – the charge had been rejected.

Yesterday I was hosting two different clients in the Accenture Liquid Studio and drove my car, not knowing how late I would be coming home. In addition to my backpack, I carried a package for UPS and a reusable grocery bag full of frozen lunches for next week.

With my hands full, I pulled my badge out of my pocket, and I suspect that my corporate AMEX card fell out then. It just so happens the Lululemon is immediately outside the lobby of my building

There are two unrelated yet interesting things about working in a mall:

  1. You’d think with stores everywhere I would get some much-needed shopping in. Hasn’t happened yet.
  2. Last week they were playing “My Heart Will Go On” from Titanic over the speakers. Never let go, Jack!

With US credit card rules being the way they are, I would not have been liable for the charge even if it had gone through. I’m glad, however, that it didn’t.

Isn’t it amazing that American Express could do that?

They know my buying habits. They know the buying habits of other people in my company. The can use the data they have to predict when something doesn’t “feel” right. We take a lot for granted these days – do yourself a favor and watch Louis CK again – Everything is Amazing Right Now and Nobody’s Happy

Artificial Intelligence (AI) is an umbrella term that describes a ton of technology – natural language processing, search algorithms, image recognition, analytics, etc. If you were to take all the technology that is lumped into AI and try to categorize them, your categorization might look like this:

Sense – perceive the world (e.g. computer vision, audio processing, sensor processing)

Comprehend – analyze and understand the information collected (e.g. natural language processing, knowledge representation)

Act – make informed decisions in the physical world (e.g. physical machine/environment control, inference engines, predictions, expert systems)

Learn – improve quality, consistency, and accuracy (e.g. machine learning, deep learning)

In the words of Ken Jennings – I for one welcome our new computer overlords.

PS – Speaking of computer overlords, last fall I got to meet Dr. Chung-Sheng Li, who played a key role in the IBM Watson Jeopardy work. Brilliant guy.

You Might Also Like