Artificial Intelligence and Machine Learning Adoption by Industry

Machine Learning Adoption by Industry: A Q&A with Stephen Woodard

As artificial intelligence and machine learning become a greater part of the business landscape, it’s interesting to see how different vertical industries are putting it to use.

How does healthcare compare to the technology sector? Do geography and location play a role? To answer some of these questions, I caught up with our own Stephen Woodard from our business development team, who looks after our customers primarily in the northeast United States.


JR: Thanks for your time today Stephen. Could you start by telling us – in broad terms – what sort of conversations you’re having with different organizations about artificial intelligence and machine learning? How are senior leaders thinking about these exciting technologies?

SW: There are definitely some interesting examples of machine learning activity taking place in different sectors, which I’ll get to in a moment. But more broadly, I want to say there are two main observations from my experience in talking to a variety of organizations.

First, organizations on the East Coast have been looking at the artificial intelligence space and the advancements on the West Coast in places like Silicon Valley. They see how that investment is accelerating, and they have recognized the opportunities these technologies can bring in terms of delivering competitive benefits. So today, we’re witnessing organizations on the East Coast begin to invest more in machine learning as they construct business strategies and tactics.

The second general observation is around a current shortage of skilled machine learning professionals – resources, people, and experience – by industry on the East Coast. I think this is understandable, as this wave of activity has only recently begun outside of the tech sector. Here’s an example: A northeastern U.S. bank does have experience using machine learning to create an eCommerce application, but does not yet have experience using it for finance industry-specific purposes. Over the next year or two, I believe this will change.

 

JR: What does that mean for the AI industry across different sectors? Is there a transformation going on?

SW: Yes there is, exactly right. And it says a lot about how exciting the field of machine learning is: we are literally on the cusp of organizations making greater strides in the AI space. In the last couple of years the penny has dropped, so to speak. Companies are now putting artificial intelligence and machine learning strategies together, looking at how machine learning can help them drive efficiencies, get to market faster, and service customers better.

In some cases, organizations are hiring teams of data scientists before they have even put their AI strategy together! That’s how much of an imperative it’s considered to be. The well-known consulting firms, also, are looking to partner with AI and machine learning leaders as they seek to grow their practices and help corporations craft strategies and commence projects.

 

JR: Great, so now that you’ve given us a broad look, can you dive into any industry-specific examples about how organizations are using artificial intelligence and machine learning right now – or at least planning to?

SW: Sure thing. Over the last decade or so, the industry has seen a lot of data science, language, and technology expertise being applied in the areas of web search, eCommerce, and social media to improve the customer experience… But now we’re seeing a shift where non-technology companies are using artificial intelligence and machine learning to improve their product or service. Specifically, we’re now seeing AI planning and activity across a broader range of industries such as healthcare, finance, and transportation.

One insurance company we’re talking to is looking to extrapolate information from their existing records – in this case stored paper files for life insurance. This business is feeding machines huge amounts of data to help the company make quicker decisions via predictive analytics. They’re looking to provide faster insurance quotes and proposals to effectively beat their competition to market. In this example, the data, via machine learning, will help them speed up the pricing component immensely, and give them a competitive edge. So what we see here is data science giving the company better accuracy and more confidence for a positive commercial outcome.

The pharmaceutical industry is a slightly different story. While the insurance sector may have its own data, pharmaceutical companies buy data, more so than creating it. Some of this data is publicly available, but in many cases, they’re buying it from hospitals or from the contract research organizations that conduct clinical trials and test the viability of various drugs or compounds. This data is then applied through machine learning again to provide predictive analytics which are enormously important in helping the pharma companies decide whether or not to pursue a drug, whether to make that further investment, and whether it is likely to pass the Food and Drug Administration (FDA). In this case, machine learning is helping the company make or save billions of dollars in R&D or product launch costs, so it’s a critical example of data science impacting the commercial reality quite substantially.

And over in the financial services industry, what we’ve noted is the companies with a business-to-consumer (B2C) unit seem to be more advanced with artificial intelligence and machine learning than those which are purely business-to-business (B2B). A good public example of this is Bank of America’s virtual assistant, Erica, which helps customers who have questions about banking products. Perhaps B2C is more visible in terms of branding and PR, or it could well be the shortest path to success. In any case, that is the situation today. But we will see, I believe, a shift where B2B financial services organizations will soon increase their use of machine learning. One example of this will be to interpret unstructured data from sources like social media instantaneously to make sure their clients are continuously well informed about market sentiment, reactions to products, and industry trends.

 

JR: We know that machine learning needs lots of training data. How are you advising clients as they put their AI and data strategies together?

SW: The most important element is quality – so we talk a lot about getting access to the right, high-quality data. Quantity of data is very important too, but quality is just such a huge requirement. Some organizations see the value of high-quality data more than others, and some almost have to go through a learning curve of first using low-quality data, and seeing a project not quite live up to expectations, before making a decision to pursue a high-quality machine learning data strategy. Which is fine and understandable – everyone is learning and growing in this field all the time.

Overall, when we’re advising clients, it comes down to looking deeply and analyzing the business case, then assessing budget, experience, and the willingness to experiment and test. Appen, of course, is well positioned to help our customers put good strategies in place. We can help organizations secure the right kind of data which is needed to create or improve products and services that rely on machine learning.

 

JR: Well that seems like a good spot to wrap it up Steve! Thanks so much for taking the time to talk with us today I really appreciate your insights from talking with a range of organizations.

SW: Anytime Jodie, thank you.

Leave a Reply

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