In the latest installment of our BrightInsight Digital Health C-Suite Series, we spoke with Ashish Agrawal, CTO at BrightInsight, about the distinction between AI and generative AI, the changes it’s driving across multiple fronts, and how BrightInsight approaches GenAI. Here’s a recap of the discussion. You can watch my interview with Ashish here, along with other interviews in the 2024 JPM Digital Health C-suite series here.
Generally speaking, artificial intelligence involves a computer or system that performs tasks a human being would normally do. Think of the voice recognition capability that’s become nearly ubiquitous in daily life. “We all use Siri, Alexa, Google voice assistant,” Ashish shares.
“The concept of a system able to consume data and make sense of it, analyze it, predict it, and even act on it, is increasing day by day.” That’s the core idea behind generative AI, which creates new content based on all the intelligence it’s gathered. For example, auto-fill features in apps are a very basic form of intelligence, “but now GenAI has taken it to a different dimension, not just completing those words, but also producing literally a new document, PPT, blog post, proposals, images & any content you can imagine across the entire spectrum.”
Over the course of his career, Ashish has seen a variety of different GenAI use cases that had a pronounced impact. We dug into these during our interview:
At BrightInsight, Ashish and his team look at generative AI in two ways. One is how it can be used internally, and the other is how it can be embedded in our products to better serve customers, patients, and caregivers.
Internally, the BrightInsight team has embarked on the journey of using AI across functions. One is collating all platform documentation so that it's easy for the ecosystem to understand, learn and create apps, and to enhance customer support. Another is using generative AI techniques to generate starter code and unit tests, and to take over some of the mundane development tasks that hinder productivity. Our devops team is looking at leveraging GenAI for better insights & expedited troubleshooting of production systems.
“From a product perspective, we've been thinking about how to enrich the patient experience.” For example, providing patient interactions based not only on their treatment stage, but on how and what they’re feeling. Or providing highly contextualized reminders to reinforce adherence, and then using that data to generate insights on how to further improve adherence. Similarly, the team is focused on “how to make life easier for the physicians,” by providing them with a summarization & concise view of patients symptoms, condition, medication and past history, thereby building capabilities that streamline workflows for care providers.
Another aspect of the company’s work with Gen AI involves the BrightInsight Ecosystem. For example, we integrate nicely with Google Vertex, exploring the medical search and other healthcare models like MedLM. Another example is Woebot Health and the “empathy engine” it developed, using cognitive behavioral therapy techniques, to help patients deal with some of the psychological aspects of their treatment, such as needle anxiety. “We continue to work with our customers, our ecosystem partners, and a lot of forward-looking folks to improve the disease management journey.”