Project

Suguard: AI-Based Mobile App for Effective Diabetes Management

Client

DiabetesLab

200+ interviews with patients

200+ interviews with patients

600 tests performed

600 tests performed

90% of cases in which the algorithm differs by no more than 20%

90% of cases in which the algorithm differs by no more than 20%

The Client

DiabetesLab is a Polish MedTech start-up founded in 2016.

The company offers an application that supports people with type 1 diabetes, helping them maintain stable blood glucose levels. The app specifically targets people who lead an active lifestyle — but it’s equally useful for any person with type 1 diabetes who wants to improve their health and well-being.

DLabs worked as a technical partner, taking responsibility for designing and developing the application up to the MVP stage in preparation for beta testing.

Project

The Problem

Studies estimate that 415 million people worldwide have diabetes. And experts predict that the number will increase to roughly 642 million by 2040.

In Poland alone, 2.9 million people have diabetes (type I and II), with one person dying every six seconds because of the disease: that’s roughly five million people-a-year — more than HIV, tuberculosis, and malaria, combined.

It’s very costly to treat diabetes. In 2018, the Polish National Health Fund spent 1.45 billion PLN delivering medication and testing slips to patients. At the same time, patients had to cover the additional treatment costs of 442 million PLN using personal funds.

A diabetes diagnosis is a life-changing moment for anyone, not least for those who are physically active. Diabetes has a drastic influence on lifestyle, and for anyone who’s newly diagnosed, it can be hard to adjust, confusing to plan a new diet, and stressful to manage the illness.

Przemysław Majewski, CEO of DLabs, is one such person struggling with diabetes. But he wanted to help other people with the disease manage the illness, plan a more balanced diet, and create personal training plans.

That’s why he created Suguard: the smartphone application that helps people with diabetes rediscover a balanced lifestyle.

Project

The solution

Suguard is a comprehensive mobile application that uses state-of-the-art AI to personalize diabetic therapy and help patients in their everyday management of this chronic disease. It aims to predict and recommend actions to stabilize blood glucose levels, decreasing the risk of complications, thus reducing treatment costs.

The solution most benefits people who have type 1 diabetes: an illness that requires more intensive treatment and an exceptional level of personal care — in particular, during high-intensity physical activity.

Quick and easy to use

The smartphone app uses familiar UI/UX and technologies the DLabs team knows well: the JavaScript language within a React Native framework ensures a high-quality, highly-portable solution.

Continuous improvements

The app gets better every day, both technically and functionally. As people use it, the growing data-set helps the AI enhance its prediction capability and make more personalized recommendations. A handful of clients use Suguard daily — with every entry improving results for all users.

Highly effective

Suguard offers 90% efficiency in predicting actions that effectively stabilize blood glucose levels, decreasing the risk of complications while helping users choose the right foods, activities, and medicines.

Award-winning

Both the concept and the application have been recognized through media coverage and awards. One prize included one-year’s funding via the TechPeaks program: The People Accelerator in Italy.

Project

What’s next?

The next step involves adapting the solution to the needs of patients with type 2 diabetes, alongside considering other potential illnesses. There are also plans to extend the AI solution to allow users to add natural inputs: for example, using voice commands.

  

Technologies used

Python

Django

pandas logo

Pandas

React Native

Java Script

Matplotlib

FireBase

Matlab

NeuroLab

SciPy

Cloud Functions

NumPy

  

The path to success

Step 1: Connecting data sets across devices

We collated data from different sources, analyzing them to give a holistic view of the patient’s current state of health

The precise detail supports more reliable predictions, helping Suguard recommend beneficial courses of action

Step 2: Glucose prediction

We designed an AI engine to analyze patient data and predict the level of glucose in the blood

The engine works with 90% efficiency regarding glucose level prediction

Step 3: Automate notifications and alerts

We implemented an automated notification and alerts system as part of the glucose monitoring

Notifications flag upcoming changes based on user activity

Alerts help users decide on physical activity, meals, insulin doses, etc

See it in action

See it in action
  

CLIENTS OPINION

format_quote

I gave five stars, in my opinion, the application will increasingly work better and better. Bluetooth will be working with for example Control plus One (…)

Client's opinion from Google Play

format_quote

The big advantage of the application is adding everything at one screen – fast and easy. Even if other applications are visually more attractive (like mySugar or Control Diabetes) you need much more time to fill them (…).

Client's opinion from Google Play

AI SOLUTIONS WE’RE PROUD OF

See other AI projects that have helped our clients achieve their business goals.

Read more call_made An AI-Based Virtual Assistant That Achieves ~50% Reduction in Meeting Organization Time

An AI-Based Virtual Assistant That Achieves ~50% Reduction in Meeting Organization Time