Skip to main content

7 AI/ML Use Cases to Watch

Dan O'Keefe, Appian
March 14, 2023

2023 is looking likely to be a breakout year for artificial intelligence (AI) and machine learning (ML). Some industry-watchers predict that recent breakthroughs in AI might lead to a new revolution in society akin to the industrial revolution, the invention of the internet, or the advent of the smartphone. Yet, 2023 doesn’t mark the invention of AI—just the year it went viral thanks to OpenAI’s ChatGPT technology.

A wide variety of Industries have been using AI for decades, which brings up the question: What are some common AI/ML use cases? What use cases should business and IT leaders track, as they shape a holistic process automation strategy?

Why is AI/ML important?

Before we go further, let’s define some of these terms. Artificial intelligence refers to computer systems that mimic human thought and decision making. Machine learning algorithms use mathematical formulas to learn from data sets and perform tasks better over time.

While generative AI tool ChatGPT has received much buzz this year, it’s far from the first widespread AI use case. For example, finance and investment trading organizations were early adopters of AI/ML capabilities, using AI/ML to decide when to buy or dump assets on the stock market and other exchanges (otherwise known as high-frequency trading). Other industries have also made widespread use of AI—from the healthcare sector to manufacturing to utilities—and we expect that adoption will only continue to grow. 

8 experts weigh in on the future of AI. Download their insights in the 2024 AI Outlook.  

7 AI/ML use cases to watch.

Let’s examine a few compelling use cases that continue to develop. 

1. Document processing.

Most organizations are drowning in documents. Whether it’s paper checks, electronic invoices, or bar code scans, companies often spend a lot of time processing documents. Intelligent document processing (IDP) allows companies to pull data from documents at scale without significant manual labor. This saves organizations massive amounts of time and effort—and money. Plus, it reduces the potential for human error inherent in typing documents manually.

2. Financial fraud monitoring.

One of the most widespread AI/ML concepts involves anomaly detection within data sets. When you train AI on a data set, the AI can develop a baseline of behaviors. Whenever something far outside the norm—an anomaly—occurs, the system can flag this anomaly for further analysis. Financial institutions use anomaly detection, during the crucial Know Your Customer (KYC) process to flag transactions for fraud or identity theft. For example, if someone’s credit card shows a large purchase made in a different country at the same time as a purchase in their hometown, the company could lock the card and run the purchase by the cardholder.

3. Credit and loan risk approval.

Institutions in the finance sector can make extensive use of AI/ML capabilities to better understand creditworthiness and money lending risks. AI can analyze multiple data points like credit histories, credit utilization, and financial statements to understand whether it’s a safe bet to shell out credit to a potential borrower for a credit card, mortgage, or business loan. Plus, over time, AI can analyze data to spot potential patterns among default risks that human analysts might not see otherwise. This not only reduces risk but also makes processing applications far more efficient for loan officers and underwriters. 

What’s next in compliance for financial institutions and how does AI/ML fit in? Read also: The Future of KYC Compliance: 4 Trends to Watch.  

4. Medical imaging and diagnostics.

One exciting area where we’ll see more use of AI/ML is the healthcare sector. For example, medical imaging professionals like radiographers and ultrasonographers will use more artificial intelligence to note potential issues in patients’ medical scans. Using machines to find potential issues and having a technician interpret the findings can be more accurate than relying on the erring human eye. Additionally, physicians will use patient input for symptoms that allow AI to diagnose diseases and even potentially recommend treatments.

5. Customer service.

One of the most interesting AI/ML use cases comes in the form of customer service. Chatbots can use natural language processing (NLP) to respond to customer requests and kick off other workflows on the backend to help solve a customer’s issues. For example, the chatbot might process low-level refunds or craft email responses for human service agents to review. This saves service reps significant time in their day-to-day operations and improves the customer experience, allowing companies to improve long-term customer satisfaction.

6. Energy prediction.

For decades, utilities have worked on smart grids. Devices like smart meters help utilities plan for peak usage times. AI and machine learning can analyze historical data like energy usage, weather patterns, and other variables to better forecast demand. This helps utility operators predict energy usage and increase supply when demand grows without overloading the system (or helps them prepare for outages). 

Plus, this new technology plays a major role in sustainability efforts, as AI can help optimize energy efficiency, usage, and distribution patterns and prevent waste. This also helps reduce costs for both utilities and industrial and residential customers.

7. Supply chain management.

The past few years have shown the fragility of global supply chains. The COVID-19 pandemic upturned typical supply/demand, leading to peaks in some areas and valleys in others. It’s hard to predict demand based on historical data alone—modern supply chain leaders need to use more sophisticated forecasting methods for supply chain management.

AI/ML capabilities let supply chain professionals better predict demand with real-time data across multiple data points to prevent shortfalls. They can also use AI to help with tasks like pricing, predicting weather patterns and routes for ships and transports, and building more responsive supply chain networks with their vendors and partners.

Deploying AI doesn't have to come at the cost of data privacy. Learn why with the eBook: Implementing Private AI: A Practical Guide.  

AI/ML capabilities keep evolving.

Want to learn more about how AI/ML fits into your overall automation strategy and what’s on the horizon for these technologies? To delve into emerging data and automation trends, get the Gartner® Emerging Tech Impact Radar: Hyperautomation.