Dr. Yossi Bahagon, is the Co-Founder & Chairman of Sweetch an Artificial Intelligence based platform for large scale prediction, prevention, management and outcome improvement of chronic disease.
Could you discuss your experience of being a family physician and how it led to the creation of Sweetch?
Working as a family physician and a medical team member allowed me to see things from my patients’ point of view, understand their needs and the challenges they face and, ultimately, acknowledge the gaps in the healthcare system. One of the main challenges that I witnessed was the lack of patient adherence. My patients were struggling to keep up with their healthcare regimens, which led me to notice a missing element between medical teams and patients. Even if you prescribe the most accurate and effective intervention for your patient, it simply won’t matter if they don’t follow your recommendation. There can be billions of dollars and ten years of research invested in a certain treatment, but it will never work if the patient doesn’t comply. The need for a system to help patients adhere to their treatment regimens became very evident. The ability to significantly impact the healthcare system is dependent on the most underutilized “resource” – the patient. We invest in hospital equipment, in various advanced medical devices and treatments, but at the end of the day it all comes down to the patient’s compliance. This is why Sweetch’s potential is so significant – it sheds light on the most important issue at hand, supporting patients throughout their journey every step of the way, offering a trusted voice that understands their needs and what drives them to action.
Sweetch is about precision engagement, could you define this term and explain why it is so important?
The medical industry is moving from a one-size-fits-all system to more personalized approaches and therapies. Precision medicine now considers the patient in a holistic way – their genetic profile, co-morbidities and more, with the aim of providing the most precise treatment plan for that specific individual. What Sweetch brings to the table is holistic “precision engagement.” We use the same tailored approach to serve patients in a way that can adapt to their unique life-print (like a fingerprint). Sweetch touches patients at the right time, tone and place, initiating engagement to prompt them to action not just in the short term, but also sustaining long-term compliance. Even if we succeed in changing 10% of patient behavior at scale, this could have a huge impact on patients’ lives and allow the healthcare system to maintain quality service alongside the digitalization and hyper-personalization movement we are seeing. A patient is not a “walking disease”, hence, to improve outcomes we must look at the human as a whole.
What type of data does Sweetch collect from a user’s smartphones and other connected devices?
Sweetch collects various types of data from the user’s smartphone and other connected devices. This includes raw data based on their activity including location, Wi-Fi connection, sleep patterns, weight, glucose levels, and more. Unlike Facebook and Google, we don’t use this data to promote customized advertisements. Instead, we use it to promote our users’ health, prevent disease and improve their quality of life in a hyper-personalized way. So, we’re using the same tools, but for a completely different objective. What’s important to note here is that all the data being collected falls under required regulations, including HIPAA and GDPR compliance.
Could you discuss how the Sweetch system then creates a picture of what the patient’s life looks like?
Our technology processes and analyzes the raw data to understand the user’s behaviors and habits. For example, if the user is connected to a certain Wi-Fi router at a particular time on specific days, we can extract what this place means to the user. This allows Sweetch to know when, where and how to intervene.
How do you trigger behavior based on the specific patients’ life habits and personality?
As humans, we have repeatable and predictable patterns. The more data we collect, the more accurate we can be in understanding patients’ lives to engage with them in a way that will encourage action. Based on Sweetch’s ability to identify these behavioral patterns and habits, we can estimate where and when users will most likely be motivated to action and comply. For example, based on the data collected, Sweetch will predict what time they’ll be heading to sleep, sending them a reminder right before bedtime to check their glucose levels or take their medication. The ability to have insights into a patient’s life habits is essential for Sweetch to know when and where to intervene, and, not less important, also when and where not to. Sweetch will know when the user is most likely to ignore the messages – e.g., when they’re in a meeting, driving etc. Not only would the message be ineffective, but could also trigger frustration, leading to non-adherence to their required treatment or healthcare regime.
A lot of emphasis is placed on meaningful activity such as duration and intensity of the activity. How do you ensure that a user does not get overwhelmed, lose hope, etc?
This directly relates to the adaptiveness of the technology – Sweetch puts emphasis on individualization, hyper-personalization and sustainability. As each person has individualized behavioral patterns, there is no specific timeline or target to reach the ultimate objectives. The slope of compliance is determined by the patients themselves, and their own personal journey to achieving their goals. This promises sustainability and long-term compliance. What is unique about Sweetch is that the system completely adapts itself to the individual’s behavior and, as a result, avoids unachievable goals and expectations. So, if the system learns that the user isn’t on the right track to succeed, instead of pushing them to do more, it automatically adapts the prompts and lowers the level of intensity to match the new goals, thus encouraging the user to keep going. This is crucial for maintaining trust, as well as for the continuation of the patient’s behavioral change in the long run.
Could you discuss what is Just-in-time-Adaptive-intervention (JITAI) and how Sweetch applies this technology?
We are the first behavioral science company to leverage AI & EI (emotional intelligence) through our digital therapeutic solutions. Our proprietary JITAI (just-in-time adaptive intervention) engine converts millions of data points from the user’s smartphone, other connected devices and continuous interactions with Sweetch into fully automated, hyper-personalized recommendations that automatically adapt to every individual’s personality and behavior in real-world context. As we mentioned, the system is consistently adapting to the user’s life habits, ensuring that the messages received are in the right time, place, tone of voice and context. This is how Sweetch motivates each individual to adhere to his or her health program and achieve personalized micro-goals and long-lasting health improvements. Our JITAI technology can be used across multiple chronic conditions, including cardio-metabolic diseases, autoimmune disorders and oncology.
What are the different intervention modules that are offered?
One of the things that makes Sweetch unique is that we don’t see the patient as a walking disease, but as a human being with, among other things in life, a mission to cope with a health condition that requires attention on an ongoing, daily basis.
Sweetch understands that the user’s feelings matter, and that those feelings have a significant role to play in the patient’s journey. The platform offers an all-encompassing view to ensure that the healthcare providers look at the patient as a whole. We touch on lifestyle medicine aspects (activity, weight management, etc.) as well as clinical aspects, such as disease management, patient reported outcomes and more. Every aspect the system needs to consider is covered by Sweetch and balanced in a way that encourages the patients to move to action, without overwhelming them with too many prompts a day.
Can you share with us insight on who provides the care packages that the AI solutions are based on?
We leverage two main resources for insights. The first is well-established guidelines. For example, our diabetes prevention program is based on CDC guidelines – pre-established and followed throughout the entire patient engagement process.
The second is our partners and customers, with whom we work very closely. We encourage them to take their own content and plug it into the system in a way that is highly user-friendly, mobile-based and visually appealing.
Thank you for the great interview, readers who wish to learn more should visit Sweetch.
- GM Announces $300 Million Investment in Chinese Startup Momenta
- Thoughtful Automation Raises $5 Million Seed Round
- Minority Voices ‘Filtered’ Out of Google Natural Language Processing Models
- Blair Newman, CTO of Neuton – Interview Series
- Adobe Makes Key New Announcements Regarding AI-Powered Capabilities