Sunday, January 23, 2022
HomeRoboticsTerry Poon, Co-Founder & CTO of Twin Well being - Interview Sequence

Terry Poon, Co-Founder & CTO of Twin Well being – Interview Sequence


Terry Poon co-founded Twin Health and drives the vision for technology development. He leads the development of Twin’s innovative platform and unique algorithms to improve human metabolic health using IoT, machine learning, and Digital Twin technologies. Prior to Twin, Terry was VP of Engineering at Jasper Technologies, where he built global software engineering teams and served as a lead architect of the company’s IoT cloud platform. In addition, Terry led engineering efforts for Jasper’s launch in the China market, the fastest-growing in the company’s history. Prior to Jasper, Terry held engineering and management positions at Oracle.

What initially got you interested in computer science and eventually IoT?

I started programming on an Apple IIe at a young age and was absolutely fascinated by the magic of computers to transform ideas into reality with just a few keystrokes.

I continued to pursue this passion at MIT, graduating with Bachelor’s and Master’s in Computer Science. I wrote my Master’s thesis on the topic of Software Agents, autonomous software programs that use AI to reason about the world, make decisions, and take actions on your behalf.

After graduation, I worked at Oracle for a few years, then joined an early-stage IoT startup called Jasper Technologies that had a vision to merge the digital and physical worlds by creating a platform to connect and manage diverse devices across global wireless networks. Over about 10 years, we scaled from 0 to more than 160 million devices running on the Jasper platform around the world.

Could you share the genesis story behind Twin Health?

Twin Health invented the Whole Body Digital Twin™ to help people reverse and prevent chronic metabolic diseases.  We recognized that the magnitude of these diseases was increasing around the world, yet no one had solved the root cause of these diseases.  We also recognized that the underlying challenge with such conditions was one of the most complex systems — human metabolism, which differs at the individual level. We felt uniquely positioned to use our experience with AI and IoT technology, and combine with deep medical science to provide a solution for people.

Part of our inspiration came from Jasper Technologies. While at Jasper, we collaborated with many innovative companies around the world, including Tesla. Tesla builds a Digital Twin software model of every car it sells; Jasper built the IoT platform that enables Tesla’s infrastructure to diagnose issues and update the car’s software automatically, keeping each car in top condition at all times.

Cisco Systems acquired Jasper in 2016. In 2018, Jasper CEO Jahangir Mohammed joined with myself and our third co-founder Maluk Mohamed to start Twin Health in order to apply this Digital Twin concept to the human body. At Twin, we have invented our proprietary technology called the Whole Body Digital Twin™, which combines IoT Sensors, Machine Learning, and Medical Science to reverse chronic metabolic diseases.

For readers who are unfamiliar with this term, what are some examples of chronic metabolic diseases?

Chronic metabolic diseases refer to conditions related to a disrupted metabolism, such as diabetes, prediabetes, hypertension (high blood pressure), dyslipidemia (high cholesterol), obesity, and fatty liver disease.

What precisely is the Whole Body Digital Twin™?

The Whole Body Digital Twin™ is a dynamic, digital representation of each person’s unique metabolism built from 3000+ data points collected daily from sensors, combined with Machine Learning models to determine cause and effect relationships, predict future metabolic states and recommend specific actions to each individual to improve their own health.

Could you discuss what some of the non-invasive wearable sensors are and what data points are collected?

The sensors include a Continuous Glucose Monitor, an Activity Tracker, a Blood Pressure Meter, and a Body Composition Scale. Using these sensors, we collect more than 3,000 data points per person per day, including key metabolic health signals like blood glucose, per-minute heart rate, heart rate variability, blood oxygen, steps, sleep stages, systolic pressure, diastolic pressure, weight, visceral fat, etc. With these signals, we are able to gain deep insights into your metabolism, i.e. how your body responds as you go through everyday activities like eating different foods, doing different kinds of exercises, sleeping, etc.

What does the machine learning system analyze and look for with this collected data?

Our Data Science team has developed multiple Machine Learning models to predict various biomarkers and recommend specific actions to improve those biomarkers.

For example, metabolic responses vary dramatically from person-to-person, and for each person from food-to-food and from time-to-time; two people can eat exactly the same meal and yet have completely different responses. Our models use more than 100 ML features to predict the responses accurately for any given meal and any given person.

How is this information then shared with the end user?

Our Whole Body Digital Twin™ provides clear health guidance to our members in two ways. First, it automatically analyzes your complex biomarkers everyday to generate a simple visualization of your current metabolic health, including an overall score and drilldowns into every facet of your metabolic health. Second, it generates highly personalized recommendations to improve your health by applying factors like Precision Nutrition, Activity, Sleep, and Breathing.

Could you discuss some of the results from different trials that have been undertaken?

Twin Health’s clinical research team has been conducting a Randomized Controlled Trial for more than 1 year and has published our results in multiple peer-reviewed journals. This is the world’s first randomized controlled trial (RCT) for reversing chronic metabolic disease using digital twin technology. We’ve seen first of its kind health outcomes. Patients had an average HbA1c reduction of 3.1 (average baseline 8.7) with over 90% achieving type 2 diabetes reversal (HbA1c less than 6.5), and 92% eliminating all diabetes medications including insulin. Furthermore, patients had an average weight reduction of 9.1 kg (20 lbs). Among patients with baseline abnormal liver function (as defined by an elevated ALT on clinical lab values) there was an average ALT reduction of 24 units/L. In comparison, the control group did not achieve diabetes reversal, any reduction in diabetes medications, or a significant improvement in weight or liver function.  The scientific outcomes from the RCT and the commercial results have been published in 21 peer-reviewed international medical journals and conference papers and abstracts.

What is your vision for the future of personalized healthcare and medicine?

Human beings are incredibly diverse; we all have unique biology, different metabolic responses, and different lifestyle preferences for foods, activities, etc. Because of this, a one-size-fits-all solution doesn’t work; healthcare should be precise and tailored to the needs and preferences of every individual.

At Twin, we believe in the power of truly personalized Precision Health to detect and reverse diseases in the most effective way for each person, keeping everyone healthy for life.

Is there anything else that you would like to share about either Twin Health or the Whole Body Digital Twin™?

We live in an age of rapid innovation driven by AI, IoT, and other disruptive technologies. At Twin, we are committed to AI for Good — using AI to understand the most complex system of all, the human metabolism, at the individual level. We are applying these technologies to solve the massive challenge posed by chronic metabolic diseases, which affect more than 1 billion people around the world. Please see twinhealth.com to find out more.

RELATED ARTICLES

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Most Popular

Recent Comments