The Emerging Cognitive Enterprise

Like so many are doing now, this past holiday season, I ordered a mattress online. It was supposed to arrive on a Friday. On Friday, I got a notification that it would arrive on Saturday. On Saturday, the delivery got pushed to Tuesday. What a mess.

But was this really an unpredictable situation? Every year, online retailers and shipping companies plan for the holiday season using analytics software. But modern artificial intelligence (AI) methods enable us to predict delivery times more accurately and enable the shipping company to plan for pending deliveries with greater precision.

Modern AI Methods

In the past few years, researchers have invented new AI methods, such as deep learning and reinforcement learning, that have revolutionized the field. Unlike past analytics methods that depended on statistical analysis and rule-based systems, modern machine and deep learning methods learn from data or "experience," such as the piles of data that retailers have about every holiday buying season.

These modern AI methods will change every aspect of how we do business: screening and matching candidates to jobs, targeting marketing campaigns to clients, scoring sales leads and forecasting revenue, optimizing supply chains, and many more use cases.

Steps for AI Success

So how do you get started?

1. Identify a low-hanging use case that has meaningful impact for your business. This could be as simple as optimizing how much cash a retail bank branch keeps on hand based on historical data or how many people service a retail store at different times during the day.
2. Figure out your data strategy. Do you have the data to actually be able to build a machine learning AI model? The test here is whether by simplifying your use case, you can use historical data to manually forecast the future. Can you then scale this to handle the complete complex use case?
3. Where possible, use prebuilt AI models and consider hiring an AI consulting team to help you get started. These will accelerate your AI journey.

The most important consideration when getting started with AI is making sure that there is clear business return on investment. Start small, with a clearly defined use case that can demonstrate benefit. Have an AI win and then you can start expanding to more use cases.

I hope you enjoy this edition as we explore how to get started with AI and transform your business.