Adopting AI in the enterprise: General Electric

Karley Yoder on what GE Healthcare has learned as it embraces artificial intelligence.

By Medha Agarwal
June 15, 2017
Kaleidoscope. Kaleidoscope. (source: Rudolf Ammann on Flickr)

As artificial intelligence technology has advanced, nearly every company has begun to respond to the promise—or threat—of AI in its industry. This post is the start of a series of interviews with executives from companies outside of the traditional boundaries of Silicon Valley. We’ll talk to them about the ways they’re approaching AI strategies and how they are leveraging the latest approaches in machine learning to provide the best products for their customers.

In our first interview, we hear from Karley Yoder, senior product manager, advanced analytics at GE Healthcare. Over the last seven or so years, GE has undergone a digital transformation—repositioning itself from a traditional manufacturing company into a software, services, and manufacturing company. For companies that aim to transform themselves through software and analytics, an understanding of machine learning and artificial intelligence will be essential. Our interview has been lightly edited for clarity.

Learn faster. Dig deeper. See farther.

Join the O'Reilly online learning platform. Get a free trial today and find answers on the fly, or master something new and useful.

Learn more

How do you leverage AI and ML to create a better product?

At GE Healthcare, we work closely and consistently with customers to understand their greatest pain points and ensure our products address those needs. It’s important to note that AI by itself is not a product but a powerful enabler to create better products. We leverage the technology to create products that benefit from the constant learning and improvement inherent in AI, turning the data into insights that have the potential to improve the quality and efficiency of care.

What steps have you needed to take in order to build a team that could grasp and apply recent advances in AI and ML?

GE has been a leader in AI for decades through the accolades and products produced by our Global Research Center. Our current team leverages these capabilities and resources, but we also aggressively recruited fresh, bright data science minds to complement this legacy knowledge. As a traditional hardware company undergoing a massive digital transformation, we have committed billions of dollars to building out our digital and AI competencies through our industrial IoT platform called Predix.

Are there other use cases within GE today?

GE is using AI to solve the world’s most pressing problems across all our industrial business lines. In health care, we are partnering with top medical institutions (UC San Francisco, Boston Children’s Hospital, Massachusetts General Hospital, and Brigham and Women’s Hospital) to create a library of deep learning algorithms that have the potential to improve quality, increase access, and reduce the cost of care around the world. The library will initially focus on diagnostic imaging applications, such as one that could identify pneumothorax—a critical condition of a collapsed lung—and prioritize that case in a clinician’s worklist, so the patient can receive more timely intervention. Over time, the library will include applications that address multiple medical specialties, including pathology and genomics, and cover numerous care areas, including oncology, cardiology, emergency medicine and women’s health.

What does GE see as its competitive advantage in these cases?

GE’s competitive advantage is that in this industrial internet world, we have both a brain and a body. What does this mean? Well, it means that in the health care setting we have both the brain (Predix and our AI Platform) and the body (our expansive and trusted install base of modalities, monitoring devices, and software solutions). Without both the body and brain, it is extremely difficult to integrate product intelligence seamlessly into a workflow or continue to refine the AI models through continuous learning/training.

Are there any areas where you’ve considered leveraging AI/ML but found that the technology isn’t ready yet?

In health care, the problem is actually flipped—we see many places where the technology is ready, but the product idea isn’t advanced enough yet. Meaning, the data science work is strong, but not enough thought has been given to workflow integration and regulatory adherence.

Is GE interested in partnering with other Silicon Valley companies and startups? What do you most want from the Valley? Are there any initiatives you’d like to see the community focus on?

We are certainly interested in partnering with other Silicon Valley companies and startups—we are constantly keeping our fingers on the pulse of innovation, both internally and externally. We are focused on partners that can help us accelerate our path to product, while adhering to the uncompromising quality associated with the GE Healthcare brand.

What’s the most promising or interesting advancement you’ve seen in AI recently, and how do you think it will impact GE?

The pace of AI advancement makes it challenging to pick a single breakthrough, but one area not to be ignored is the value creation enabled by the advance of compute power. Not only does this advancement allow for the more rapid scaling of AI and deep learning models, but it also allows these models to be deployed through the cloud, an edge device, or directly within existing products—this delivery flexibility is novel and allows us to meet our GE Healthcare customers exactly where they are, making the delivery method of AI invisible to the ultimate end-user.

Could you tell me a bit about yourself, your role at GE, and what you are working on?

I am the senior product manager of advanced analytics at GE Healthcare. In my current role, I lead the work to build our machine learning platform, which will allow GE to quickly scale machine learning and deep learning models in order to introduce intelligence into our product lines. I have the distinguished pleasure of working with world-class data scientists and software engineers to accomplish this task.

I have spent the past decade focused on the health care space: spending time at Apple Health, leading implementations of the ACA exchanges at the state level, developing BD strategies for Doctor on Demand (a telemedicine San Francisco startup), and conducting clinical/software research in the health care space at Duke (oncology drug delivery) and Stanford (MRI technology development). I majored in biomedical engineering at Duke University and earned my MBA from Harvard.

Post topics: Data-Driven Business
Post tags: Enterprise AI & ML