Photo by Mohamed Nohassi on Unsplash

Artificial intelligence has been introduced to companies around the world, with some good results and some waste of resources. At this point we can see trends that will help business leaders implement worthwhile efforts. Best use cases vary from industry to industry, but commonalities abound. Businesses can benefit from looking beyond their own industry or function to see what has proved useful elsewhere. Many good use cases will work in other sectors as well.

McKinsey has recently written about nine different sectors, complementing the articles I have written on industries and business functions. Let’s extract the good ideas from across this body of work.

AI For Time-Consuming Tasks

One great example that McKinsey and I both have highlighted typifies large benefits that can be quickly implemented. Our specific case is AI-powered healthcare scribing, but managers in other industries can also benefit from the concept. Doctors and nurses use electronic medical records to document patient visits as well as to access information such as past visits and test results. Writing up visit summaries is a time-consuming and tedious task performed by high-paid workers. AI tools can listen to a conversation and prepare a summary in the appropriate format. (This works best when the doctor gets into the habit of speaking out loud the signs, such as “your lungs sound clear.” The practitioner then reviews and edits as appropriate, usually in a fraction of the time that would have been needed to write the visit summary.

Where does a company have employees spending time on tasks that an AI can quickly do? It could be sales representatives logging calls, service technicians documenting tests, compliance officers checking documents. The McKinsey writers argue for improving existing processes first, then tacking major innovations. That’s good advice.

In many of these use cases, the employees don’t enjoy the particular task. Physicians don’t like writing up patient visits, but they know they need to. Good salespeople take notes because it helps, not because they like to. So a company can gain in efficiency and also employee work satisfaction.

AI To Improve Customer Service

Help for customer service representatives cuts across several of the industries McKinsey surveyed. It’s a large, ubiquitous business function that I described as “The lowest hanging, fattest fruit in the whole orchard.” Imagine a call to a customer service representative, with an AI-augmented system listening in. The AI can pull up the customer’s history, even if the customer doesn’t know which model he owns. The AI may prompt the rep with questions to ask (“Did this problem arise suddenly or gradually?”). And when it’s helpful, the AI will pull up company policies, service manuals or trouble-shooting tips. This application is especially valuable for less experienced representatives.

AI For Compiling Information

In a number of industries, employees must pull information together from multiple sources. The McKinsey article on pharmaceuticals, for example, describes regulatory applications drawing on academic publications, databases, trial data and patents. Generative AI is great at pulling together information from diverse sources.

That concept can be applied in multitudes of cases. One application in a very different industry has been developed by WFG National Title Company (a client of mine). Closing real estate deals requires compiling multiple documents, including the purchase agreement, mortgage, various disclosures and title insurance. An AI app is fed the documents. It examines and categorizes them, checking for signatures and initials. Then the app routes the documents to an existing system to create the final signature packages. One study found that for a typical home sale, the buyers’ and sellers’ names and addresses appear 80 times on various documents. The chairman of WFG asks, “What are the odds that names and addresses are entered accurately all 80 times?” The company has found that time devoted to closings has been cut by 30 minutes on average. Multiply that savings time thousands of closings a month.

The concept could also apply to engineering designs, real estate development applications and financial risk assessments. AI can be shown the appropriate format for the final product and asked to use the various resources to write the document. It will need to be checked for errors by humans, but that is easier than writing it up by hand.

The energy and materials article mentions integrating varied data on physical assets (utility systems, machinery), such as sensors, past physical inspections and automated image capture. The end result is predicting failures and scheduling maintenance. Thinking beyond drug approval requests, the general concept is that AI right now performs well when multiple data sources must be integrated into one description or plan.

Business Applications For AI Imaging

The large language models have led to imaging innovations, with business applications starting to arrive. In consumer sales, someone might take photos of a living room and use AI to add a black leather couch to see how it would look in that particular location. Or a new home might be visualized with a prospective buyer’s own furniture. The backyard might be imagined with the maple tree having grown 20 feet higher. A kitchen window’s view could be shown at 9:00 am on a winter morning.

Retailers might record how customers walk through a store, then visualize paths with different displays and fixtures. Visualization will increasingly be used in a wide variety of applications.

AI For Consumer-Based Businesses

McKinsey’s travel article highlighted a fact important for everyone working with consumers. “… every customer gives tells. They drop digital breadcrumbs of things they like and don’t like when they bounce off of the page of a dot-com when they’re shopping; when they abandon a cart; when they return less frequently to search; when they arrive on a page only to check a single itinerary on a single day, on a single fare, rather than browsing for 20 minutes.” That has enabled websites to serve up ads to likely buyers, but it could also empower consumers themselves to use an app with access to all of their information—not just browsing history, but also credit card statements, emails, and calendar items—to get better buying options.

An AI app with this information could design a vacation consistent with my past vacations, typical price points, loyalty programs and calendar availability—in a few seconds. Although the example related to travel destinations, it could apply to any purchase, from clothing to cars to home furnishings.

Business leaders looking for opportunities to serve customers better, at lower costs, should browse widely through AI applications in a number of industries and business functions.

More By This Author:

Food Costs Are High: The Diet Of Misleading Information
Consumers Will Keep On Spending In 2024
AI Market Structure In The Future: Few Engines, Many Applications