Oracle’s GenAI strategies making impact in healthcare and beyond: Executive VP Mike Sicilia

Generative AI has become a conversation piece among our customers. We have added GenAI features in our application stack, both horizontal and vertical applications. In the latter, for example, we have seen tremendous uptake of interest in many different vertical industries, with many use cases. Last week, we announced the general availability of our GenAI-based clinical digital assistant along with an ambient listening-based technology that allows healthcare providers to automate routine tasks. There is a supply-demand problem in healthcare and a global shortage of healthcare workers. This gets particularly acute during situations like pandemics. So these innovations are game-changing and time-saving.

Also, we believe customers should be able to bring their own data to further these LLMs. But with every new technology, there are fear cycles too. We often talk about avoiding hallucinations. In other words, you don’t want the model to make something up, particularly when dealing with healthcare issues.

Also, we have been LLM agnostic and have abstraction layers on top of the LLMs our applications connect with. Over time, customers may want to influence which LLM they use based on the vertical they are in.

LLMs in general, had a breakthrough in the last 18 months and the whole category is still in the early stages. It is easy for customers to compare our generally available GenAI-powered healthcare clinical digital assistant with the existing system to measure the time saved.

We are seeing widespread adoption across many industries that consume our horizontal applications. Apart from human capital management (HCM), and job postings, we use GenAI for repetitive tasks in our accounting, and enterprise resource planning (ERP) accounting systems. Some customers are adopting AI in hospitality, restaurant businesses, and engineering construction businesses as well. The early adoption is coming from adding GenAI features into the application stack, and not by providing a toolkit to customers to say they can do whatever they’d like with it. It is much easier to consume a built-in application feature. I can’t think of any vertical industry today with no GenAI use cases. Healthcare, banking, and hospitality seemed to be early use cases for feature-based applied AI. Coming to molecular discovery and life sciences, the industry, and not just Oracle, is saying that computers would probably have an interesting application in simulating molecules, turning them into a pharmaceutical that benefits people, and simulating some adverse events.

As far as our workforce and recruiting and retaining people are concerned, we are recruiting for data science. That being said, there needs to be data and AI strategies. The fact that the Oracle database is a popular choice for the world’s mission-critical data means we have a head start in aggregating data, and helping our customers do the same. We also are hiring developers in the space as well. Quite a few system integrators are focused on change management with customers, the transformational impact AI could have on their business, retraining their employees, and providing better customer service. However because we are providing vertical edge applications and custom applications, when you apply the AI feature set into the applications, a tremendous amount of training is not needed for using them. They are features in an existing application set made more robust by the application.

Lots of regulators around the world have a lot to say about AI. It’s our job to work with our customers and be compliant. This is going to vary with the industry. While dealing with personal health information, there will be more conversations between medical providers and governments about how AI is applied. We must adjust the technology to meet those requirements. In a restaurant management business, using AI to automate menu management and other things is inherently less sensitive. We aim to make the technology easy to apply, even while dealing with sensitive and regulated industries.

So far, we have seen genuine excitement around productivity enhancement. There is a labour shortage in the world today in healthcare, and construction, for example. Our customers are not worried about losing jobs, but about making people more productive. With AI in general and generative AI, there is a new classification of data science jobs which have risen to being on par with computer engineering jobs. With any big technical revolution, there is always a new category of jobs created as a side effect.

Every vertical application of ours has a presence in India. BFSI is probably our biggest vertical and we have a highly skilled workforce in India. As far as the market itself, we have had an uptick in SaaS applications in general in India, which is new. India, because of its long history and legacy in the IT space, has many bespoke systems because it is less expensive to create what customers want. The unit economics up front may have been better for the customer, but it has led to data fragmentation and bespoke systems. We are seeing a new line of thought about getting standard systems, because, over time, bespoke systems are more important to maintain. To maximize the usage and application of GenAI applications, you need a toolkit approach to make them work with older, bespoke applications. You will see continued progress in the off-the-shelf, SaaS uptake in India. Also, Oracle has been a long-time provider of jobs and opportunities in India. Both of those things coming together will continue to have India at the forefront of technology consumption.