A continuous glucose monitor helped me discover I have hypoglycemia. Here’s how startup January uses AI to predict blood sugar spikes and crashes and coached me to better food choices.

• The healthcare startup January uses CGMs and AI to monitor glucose levels, exercise, and sleep.
• The startup has received funding from Felicis Ventures, HAND Capital, and Marc Benioff.
• I have low blood sugar, and January used machine learning to coach me to healthier habits.

The health-and-fitness-startup industry is continuing to boom thanks in part to the exploding popularity of consumer devices that can track steps, fitness, sleep, menstrual cycles, and more.

But in this growing world of health-and-wellness wearables, continuous glucose monitors, or CGMs, remain some of the most invasive and intensive options on the market. While the devices, traditionally used by diabetics, offer the ability to monitor blood sugar levels 24/7, they’re also expensive and require a prescription — plus, many require reapplying the device with a microneedle to the back of the arm every one to two weeks.

That’s where January AI comes in. The startup, which Noosheen Hashemi founded in 2017, promises a future of glucose monitoring with just one CGM. Wear it for just five days, connect a fitness tracker, and the app uses machine learning to predict blood-sugar spikes and dips depending on the food consumed, exercise levels, and time of day.

Hashemi told Insider in an interview that blood-glucose levels are an important health indicator that can help people prevent diseases later.

“I wanted to have a fuller picture of one’s health, more than just what a doctor will tell you because as we wait, the choices for treatment are fewer and further between,” she said. “Powerful monitors can improve peoples’ health. Healthcare today is focused on illness, not health, so we’re trying to focus on that.”

The startup has raised $28.3 million in funding from investors such as HAND Capital, SignalFire, AME Cloud Ventures, and Felicis Ventures in addition to well-known angel investors including Marissa Mayer and Marc Benioff.

While removable fitness trackers such as the juggernauts Fitbit, Whoop, Oura, and even the Apple Watch have long dominated the wearable world, CGMs are still a relatively small niche, with young startups such as Levels and Nutrisense letting users track their blood-sugar levels to monitor for diabetes and manage weight.

January, which came out of stealth in 2020, caught my eye specifically because it promises to provide deeper health insights without requiring users to always wear a CGM.

I tried out January’s platform for more than a month, which included five days of AI training while wearing the CGM followed by weeks of machine-learning-generated data about my food and movement habits. Spoiler alert: The startup provided me with helpful health recommendations that I’ll be incorporating into my everyday life, and it helped me identify and manage hypoglycemia.

January provided me with a CGM and access to the startup’s app to review for this story. The startup charges $288 for the first month of usage, which includes a telehealth evaluation, a starter CGM, and AI calibration. After that, the app’s membership fee is $28 per month.

After a medical professional reviewed my intake form, I received a Libre FreeStyle 2 GCM

My January journey started with an online health evaluation where I answered questions about my existing health conditions, medications, medical history, and allergies. After a physician licensed in my state reviewed my evaluation, I was approved for a CGM prescription, which was delivered to me via mail.

CGMs continually monitor blood-glucose — or blood-sugar — levels. Unlike glucometers, which test blood-sugar levels via pinpricks of blood, a CGM is often applied to the back of the arm and needs replacing every seven to 14 days, depending on the model. The Food and Drug Administration approved the first CGM in 1999, and in recent years they have become more available and less expensive — leading to increased use by consumers and healthcare startups outside of diabetes management.

Mike Snyder, January’s other cofounder and the director of genomics and personalized medicine at Stanford Medicine, told Insider that this new technology is pulling back the curtain on people’s health in a way that wasn’t possible before.

“This is a combination of pulling in new tech and using everything to your advantage,” he said. “Through monitoring, we can push diabetes off for a number of years, and those numbers are getting worse as people live longer. We’re trying to avoid people heading in that direction.”

I had never applied a CGM before, but the app seamlessly guided me through the experience

Unlike other CGM startups and wearables, a big draw for January is that you don’t have to wear one all the time and keep reapplying another every two weeks.

What’s more, the CGM doesn’t even have a “needle” — instead, it has a thin, flexible filament that goes just under the skin.

As someone with a slight fear of needles — I’m alright with shots and blood draws, but the idea of sticking myself with something makes me queasy — this was all great news. Still, I was anxious about applying the CGM and paced around the Insider office for multiple hours before working up the courage to put it on.

However, the CGM was surprisingly painless, and I appreciated that the startup’s app had step-by-step instructions — with video tutorials — for opening the packaging, assembling the applicator, and applying the patch.

It was a more in-depth setup experience than other wearables I’ve tried

In addition to using a CGM to track glucose levels, January also requires users to connect a fitness wearable to track heart-rate data. The app is currently compatible with three devices: Fitbit, Oura, and Apple Watch.

I’m a runner and wear a Coros Pace 2, so the January team provided me with a Fitbit Inspire for the review.

Because you’re connecting three different apps and two wearables — the CGM to its app and then to January, and the Fitbit to its app and then to January — the January setup period was more involved than other programs, and it also took me a few different times of logging in and out of various apps and reconnecting the devices before all the programs were talking to one another correctly. The time investment did end up paying off when I could see my heart rate and glucose data together in the January app, but users should be prepared to spend a little extra time getting everything connected properly.

My first CGM also fell off a few days into wearing it, coming loose when pulling my arm through a tight sports bra. January, which promptly sent me another patch to apply along with sensor covers that provide an extra layer of protection, said it’s rare for its patches to come off.

After the CGM and fitness tracker were synced, January kicked off five days of AI training

During the five-day AI-training period, January collected my glucose and heart-rate data around the clock. Additionally, I started tracking everything I ate and entered it into a food log within the January app.

Much like the initial-setup process, the AI training was more involved compared to other wearables I’ve tried, many of which are much more “set it and forget it.” While CGMs track your glucose levels 24/7, I had to regularly open the FreeStyle LibreLink app and hold my phone up to the patch to scan for the data it had collected since the last reading, which was then sent to January.

This step is especially important to remember after eating so the app can begin to identify how different foods affect your blood-sugar levels in real time.

The Fitbit integration, meanwhile, was good to go after the initial-setup process and didn’t require anything else for January to begin collecting my heart-rate data.

While using the CGM during the AI training process, I discovered I have low blood sugar

Early into my time with the January app, I discovered that my blood sugar was consistently on the low end of a healthy range, which, when fasting, is between approximately 70 and 100 mg/dL, according to Mount Sinai. Additionally, my blood-sugar levels would crash a few times a day, dipping into the low 60s during the day and occasionally as low as 54 mg/dL overnight, even with regular meals and snacks.

Low blood sugar, or hypoglycemia, is less common than high blood sugar, or hyperglycemia.

Though I didn’t have to worry about hyperglycemia, discovering my low blood-sugar levels helped me understand a set of symptoms I regularly experienced: lightheadedness, fatigue, getting “hangry,” and having shaky hands.

When the training period is complete, you can remove the CGM and continue to receive AI-powered glucose predictions throughout the day

At the end of January’s AI training, I unlocked a suite of tools to track my food, exercise, blood sugar, and more. The startup’s landing page includes a daily health recap with my calories and fiber consumed, whether I took post-meal walks, and how much sleep I got the previous night. A large line graph also includes a timeline of my predicted blood-glucose levels, heart rate, and time stamps for meals and exercise, giving me a big-picture look at how diet and exercise were impacting my blood-sugar levels throughout the day.

When logging my food, January provided me with predictions for how certain foods, such as a bag of pretzels, might spike my blood sugar more than others, such as a serving of green beans. These estimations are AI-generated and come from the data the CGM collected while I was wearing it and tracking my food during the training period.

January also sent me push notifications to keep me engaged with my diet and exercise goals throughout the day. The app pinged me if it was late in the afternoon and I hadn’t yet logged lunch; when I ate a meal and didn’t soon head out for a short walk; or to let me know I met my recommended fiber intake — which can help regulate blood-sugar levels.

During the period in which I was reviewing the startup, January launched an AI health coach that was embedded directly within the app. I could ask the digital assistant, “Jan,” various questions about my food choices — such as what I should have for dinner, how I can eat more fiber, or asking it for a suggested weekly meal plan — and it would provide recipes and information while taking dietary and other preferences into account.

While using the app, I learned some surprising things about my health — and how I could tweak my eating and exercise habits to avoid a midday sugar crash

Even with hypoglycemia, I learned through the January app that my breakfasts were spiking my blood sugar and leaving me hungry well before lunchtime — and sometimes craving a mid-morning snooze. I often eat lots of fruit in the morning to kick off my day with a good dose of nutrients and fiber, but January’s insights helped me realize I was neglecting protein and healthy fats that would keep me full longer and give me enough energy to make it to my next meal.

The startup also helped me stay active throughout the day as it related to my post-meal routine. As an avid runner who exercises almost every day, I had never thought about the timing of my workouts, and I can get lazy with movement throughout the day if I complete an hour-long workout in the morning. But January encouraged me to take a short walk after eating, which can help moderate glucose levels following a meal.

One of my favorite things about January is that it challenged my perceptions about “healthy” food and encouraged me to make small tweaks to my existing diet so I could continue to have food I enjoy while mitigating negative impacts on my blood sugar. As an endurance athlete, I previously didn’t think much about eating other than trying to get in three mostly-healthy meals per day, a couple of snacks, and enough fruits and vegetables. As a result, I often struggled with rapid-onset hunger that took me from zero to dizzy in a matter of minutes, and I was regularly under-fueled for workouts and sluggish after eating big meals.

But with January’s recommendations, I started making small tweaks that helped me feel more balanced with my energy throughout the day.

Overall, I really liked using January and found that its health insights motivated me to make some changes to have more consistent blood-sugar levels

While January largely caters to common hyperglycemia and an estimated 130 million diabetic and pre-diabetic Americans, I still gleaned helpful information about my own health, low blood sugar, diet, and exercise. The startup’s recommendations and health coach provided me with actionable steps to improve my eating habits help avoid dips and spikes, and I’ll be carrying these new habits on with me following this review.

However, this might not be the tech for you if you struggle with closely monitoring your food and exercise levels. Keeping a food diary is proven to help with weight loss and healthy eating, and in January’s case, it’s a crucial data point to continue receiving AI-powered predictions about your glucose levels. On the other hand, people prone to disordered eating may not do well to log every meal.

The initial setup and AI training is a bigger time and labor investment than competitors.

The payoff is very high, though, especially because the startup is harnessing AI to provide a great digital-health experience following the setup process. The CGM market is booming right now as devices become less expensive and as digital-health startups look to harness the power of blood-glucose tracking.

January is a game changer in that regard — while I was pleasantly surprised that applying the CGM didn’t hurt, I was even happier knowing I wouldn’t have to apply a new one every two weeks to keep using the app.