Prevent Blood Sugar Spikes | Personalized Software

Understanding the Rise of Personalized Nutrition with January AI

Type 2 diabetes, characterized by chronically elevated blood sugar levels, affects a significant portion of the population. Currently, approximately 9% of Americans are living with this condition, while an additional 30% are considered to be at risk of developing it.

Introducing January AI: A New Approach to Dietary Management

January AI, a four-year-old, subscription-based company, launched in November with a novel solution. They provide customers with individualized nutritional and activity recommendations. This is achieved through a combination of extensive food-related data collected over three years and a unique profile developed during an individual’s initial four days of software use.

The Importance of Personalized Responses to Food

The need for personalization stems from the fact that individuals can exhibit remarkably different reactions to the same foods, ranging from rice to salad dressings. This variability underscores the limitations of generalized dietary advice.

Cofounder and CEO Noosheen Hashemi, alongside cofounder Michael Snyder – a genetics professor at Stanford specializing in diabetes and pre-diabetes – believe this technology is both groundbreaking and potentially life-altering.

Securing Investment for Growth and Innovation

Investors share this optimistic outlook. Felicis Ventures recently led an $8.8 million seed funding round, with participation from HAND Capital and Salesforce founder Marc Benioff. Previous investors include Ame Cloud Ventures (led by Jerry Yang), SignalFire, YouTube cofounder Steve Chen, and Sunshine cofounder Marissa Mayer, bringing the total funding raised to $21 million.

A Deep Dive with the Founders: Hashemi and Snyder

We recently spoke with Hashemi and Snyder to learn more about their innovative platform. The following is an edited transcript of our conversation.

The January AI Platform: How it Works

A Multiomic Approach to Glycemic Response Prediction

TC: What is the core functionality of your platform?

NH: We’ve developed a multiomic platform that integrates data from various sources to predict individual glycemic responses. This allows users to make informed choices before consuming food. We incorporate data from heart rate monitors, continuous glucose monitors, a 1,000-person clinical study, and a comprehensive atlas of 16 million foods.

Using machine learning, we’ve derived nutritional values and created detailed nutritional labeling for these foods. The system can predict a customer’s glycemic response to any food in our database after just four days of personalized training. Users don’t even need to consume the food to understand its potential impact; our product provides that insight.

Predictive Glucose Monitoring: A Significant Advancement

TC: While glucose monitoring isn’t new, your approach is predictive. Why is this distinction important?

NH: Our goal is to restore enjoyment to eating and eliminate feelings of guilt. We can, for instance, calculate the amount of walking required after eating a specific food to maintain healthy blood sugar levels. Simply knowing what *is* happening isn’t enough; we aim to tell users what they need to *do* about it. For example, if someone is considering fried chicken and a shake, we can estimate they’d need to walk for 46 minutes to stay within a healthy range. Would they be willing to commit to that activity? If not, perhaps enjoying that meal on a Saturday would be a more suitable option.

Subscription Details and Long-Term Engagement

Pricing and Subscription Model

TC: Can you describe the subscription model and associated costs?

NH: The standard retail price is $488 for a three-month subscription, but we currently offer an introductory rate of $288.

Addressing Concerns About Subscription Attrition

TC: Are you concerned that users might gain sufficient insight from the product to modify their behavior and then cancel their subscriptions?

NH: No. Individual profiles change with pregnancy, age, and travel habits. People don’t consistently eat the same foods.

MS: I’ve personally used continuous glucose monitoring wearables for seven years, and I continue to learn new things. You might discover, for example, that white rice consistently causes a significant blood sugar spike. This is true for many individuals. We are also planning to introduce a year-long subscription option, recognizing that people may occasionally need reminders of the value these tools provide.

Practical Applications and Future Development

Real-World Scenario: Choosing a Healthier Pizza Option

TC: How does this work in a practical setting, such as when dining at a restaurant? For example, if a user is craving pizza but is unsure which option to choose?

NH: The platform allows users to compare glycemic curves for different pizza options, identifying the healthier choice. It also estimates the amount of walking needed to offset the impact of each option based on its toppings.

Data Input Methods

TC: Does the user need to manually input all the ingredients?

NH: January AI can scan barcodes, interpret photos, and accepts manual entry, as well as voice commands.

Expanding the Food Database and Data Utilization

TC: Are you exploring other applications for your extensive food database?

NH: We are committed to protecting personal information and will not sell it.

TC: Even in an aggregated, anonymized form? The database seems incredibly valuable.

MS: We are not like 23andMe; that is not our primary objective.

Surprising Insights Revealed by the Platform

TC: You mentioned that rice can significantly impact blood sugar levels. What are some other surprising discoveries users might make with your software?

NH: The variability in individual glycemic responses is remarkable, not just between different people, but even for the same person on different days. A person’s response can vary over nine consecutive days due to factors like sleep quality, fiber intake, and pre-bedtime eating habits.

Activity levels before and after eating, as well as fiber consumption, are crucial. Fiber is often overlooked in the American diet. Historically, our ancestors consumed around 150 grams of fiber daily, while the average American today consumes only 15 grams. Many health problems can be linked to this fiber deficiency.

Exploring Partnerships and Future Growth

The Role of Coaching and Integration with Existing Services

TC: Would incorporating coaching services enhance the app’s effectiveness?

NH: We don’t currently offer a coaching component, but we are actively discussing partnerships with several coaching solutions to serve as an AI partner for them.

Strategic Partnerships with Healthcare Providers and Employers

TC: Are you pursuing partnerships with healthcare companies or employers who might offer this as a benefit?

NH: We primarily sell directly to consumers, but we’ve already collaborated with a pharmaceutical company for two years. Pharma companies are interested in our platform because it allows us to use lifestyle as a biomarker. We provide them with anonymized insights into a person’s lifestyle over a specific period, helping them assess whether a therapeutic is effective regardless of lifestyle factors. This can accelerate trial phases and potentially reduce the number of subjects needed.

We are excited about the potential for partnerships with pharmaceutical companies, employers, coaching solutions, and ultimately, payers like insurance companies.