AI powered Diabetes Maintainence System

An AI-powered diabetes prediction chatbot that utilizes machine learning algorithms to predict whether a person has diabetes or not by asking 20 questions, providing personalized recommendations, alerts with 85% accuracy.

Approximately half of individuals with diabetes are undiagnosed, resulting in a significant public health concern as untreated diabetes can lead to serious complications. The problem is identifying and diagnosing these individuals in order to ensure they receive appropriate treatment and management of their condition.

Our solution is a AI-powered chatbot which will be able to predict whether a person has diabetes or not by asking 20 questions.

1. Lack of awareness: Many people may not be aware that they are at risk for diabetes and may not seek out a medical evaluation. Your chatbot aims to help users understand their risk and take proactive steps to address it.

2. Access to healthcare: By making the chatbot accessible through social media, integrating with medical and hospital  website and/or mobile app, it can help users assess their risk for diabetes without having to schedule an appointment with a healthcare provider. Moreover, we will be launching our chatbot in bi-linugal, English and Burmese.

3. Personalization: By answering 20 questions, the chatbot can provide personalized recommendations and alerts to help patients manage their diabetes and achieve optimal health outcomes.

4. Ease of use: The chatbot is designed to be user-friendly and accessible, allowing patients to easily access their data and manage their diabetes from anywhere, at any time.

5. Education: By incorporating educational resources and information about diabetes prevention and management, the chatbot helps users better understand the condition and take steps to reduce their risk.   

We will gather blood sugar levels, medication schedules and exercise habits on our chatbot which will enable us to predict the glucose level in blood as the next stage.

Our chatbot is designed to serve individuals who may be at risk of developing diabetes, as well as those who have already been diagnosed with the condition. The chatbot is able to assess an individual's risk for diabetes by asking a series of questions, and can provide personalized recommendations and alerts to help them manage their condition. By making the chatbot accessible through social media, integrating with medical and hospital  website and/or mobile apps and health care gadgets, it can reach a wide audience, including those who may not have easy access to healthcare providers.

The solution can impact the lives of those it serves in several ways:

1. Early detection: By identifying individuals at risk for diabetes, the chatbot can help them take proactive steps to address the condition before it becomes more severe.

2. Self-management: The chatbot's personalized recommendations and alerts can help individuals better manage their diabetes and achieve optimal health outcomes.

3. Convenience: The chatbot's accessibility through social media, integrating with medical/hospital websites and/or mobile apps allows individuals to easily assess their risk for diabetes and manage their condition from anywhere, at any time.

4. Cost-effective: By reducing the need for frequent visits to healthcare providers, the chatbot can help lower healthcare costs for individuals and the healthcare system overall.

Our chatbot can have a significant impact on the lives of those it serves by helping them detect and manage diabetes, and improve their overall quality of life.

1. Kaung Min Khant’s profile: Kaung’s skills and experience in the field of computer programming over 4 years, with a focus on web backend development and expertise in natural language processing and machine learning, as well as proficiency in project management tools such as Clickup, Trello, Jira, Gitlab and Github, have been served to global clients in countries such as Myanmar, Singapore, South Africa, USA, Australia, and Canada. The ability to deliver innovative solutions has been demonstrated by Kaung, who not only possesses expertise as a programmer but also an interest in project management. 

2. Yamin Thazin Oo’s profile: As a versatile and experienced product designer, Yamin excels in various roles including UX, research, prototyping, writing, and human-centered and data-driven design. She has +4 years of experience in product management and UX in the industry, including expertise in natural language processing (NLP) technology. Her NLP-based product has impacted over 70,000 people in Myanmar, showcasing her abilities in creating effective and impactful solutions. She is passionate about personal and professional growth and is highly adaptable, always open to constructive feedback. Yamin thrives in an agile, collaborative environment, with the ability to work both async and synchronously, making her an asset to any team looking to drive success and innovation in their product development, particularly in NLP-based products.

3. Min Myat Swan Pyae’s profile: Min Myat Swan Pyae (Swan) is a business strategist and tech entrepreneur in Myanmar. Swan is currently running two start-ups; Juncture, which is a human resource tech start-up and Chat Studio which is a chatbot marketing agency. Together with a strong knowledge in business operations, sales, marketing and partnership, Swan has 4 years experience of running a start-up, including fundraising and legal process. 

Kaung's skills and experience in computer programming, with a focus on web backend development, expertise in natural language processing and machine learning, as well as proficiency in project management tools, make him an ideal candidate to lead the technical development of the chatbot. Yamin's expertise in product design and natural language processing, along with her experience in product management and user experience, will ensure that the chatbot is user-friendly and intuitive. Swan, as a business strategist and tech entrepreneur, brings a wealth of experience in running startups, including business operations, sales, marketing, and partnerships, which will help in the overall success of the project, including fundraising and legal process. Together, the team has a diverse set of skills and experience that will enable them to deliver an effective and impactful solution that will improve the quality of life for individuals with diabetes and to those who are willing to prevent diabetes.

We approached in two ways for identifying the problems : Quantity Research Analysis and Human-centered Design thinking which means we observe whether the problem actually exists in reality or not. 

Based on quantity research analysis, diabetes prevalence in Myanmar is high, and relatively higher than that reported in many Association of Southeast Asian Nations countries. Based on the world bank's analysis, Diabetes prevalence is 7.1% of population ages 20 to 79.Overall prevalence of prediabetes was 19.7%: 16.5% in males and 23% in females. Diabetes is a global issue. Some 425 million people worldwide, or 8.8% of adults 20–79 years, are estimated to have diabetes. 

Based on the human-centered design, we could observe that 50% of respondents (based on 100) are having diabetes and can’t maintain the diabetes due to cost, inaccessibility and lack of knowledge using the painful device.

Our chatbot provides an innovative solution by being “ECHPAI” but not limited to. ECHPAI means effortless, cost-effective, highly accurate, personalized, accessible and integrated system with other existing systems. All these abilities are in one package by using machine learning techniques.

As we mentioned earlier, we already developed the AI algorithm for our very first MVP chatbot product. Thus, we expect to launch our chatbot product by June, 2023. Our first MVP AI chatbot can tell only whether a person has/will have diabetes.  However, we have a vision to develop an algorithm to predict the glucose level of a diabetes patient by 2024 by utilizing the data we collected from the chatbot users in our MVP product. With the help of that AI algorithm, diabetes patients will not need to prick their fingers and suffer from infinite pain frequently. Moreover, it can save costs and improve the quality of life of diabetes patients.

Our cutting-edge solution harnesses the power of machine learning to accurately predict whether a person has diabetes or not. By utilizing various regression models, we have achieved an impressive 85% accuracy rate in our predictions. This high level of accuracy is made possible by training our models on a vast dataset of over 250,000 records, with more than 20 factors taken into consideration. This extensive training ensures that our solution is able to provide highly accurate and reliable predictions, making it an invaluable tool for early detection and management of diabetes. With our solution, individuals can take proactive steps to address the condition before it becomes more severe, improving their overall quality of life.

We are still in the development stage. But, we already developed the AI algorithm to predict whether a person have or will have the diabetes. So, we believe that we will be able to finish our MVP product in 2 months and do the market testing. After 3 months of market testing, we will be able to officially launch our product in the market. According to the statistics, Myanmar has 1.9 million messenger users. As a start, we will utilize messenger chatbot as a primary platform to engage with diabetes patients. Since our core technology is able to integrate in different social media platforms, website and mobile appplication, our business model is scalable and cost-effective. Regarding our target audience in Myanmar, it is estimated that more than 2.1 million adults (aged between 20-79) have diabetes. Therefore, we aim to help and serve only 10% of diabetes patients in Myanmar in 2023 and 50% of diabetes patients in 2024.

Since our product can be accessible online, there are not many barriers to entry to the market for us. However, one significant challenge is the lack of internet and electricity access in most of the rural areas in Myanmar due to the current political conflicts. Moreover, there is also a potential risk that can happen anytime in Myanmar is that the Myanmar Military government can cut the internet across the country. In that case, we will create an application that does not require internet connection to run.

It has been only two weeks since we started working on this project. Thus, we have not partnered with other organizations.

B2B - Affiliate Business Model: Marketing and advertisments from health, pharmaceutical and lifestyle companies.

B2G - Government Procurement Model : Selling the integrated system to respective government who are willing to support quality health to the public 

Grant-based Revenue Model : Applying grants for global health care support 

Based on the impact wise, we are achieving the SDG as 

Goal 3: Good Health & Well-Being 

Goal 10: Reduced Inequalities

We will surely have the operational and development expenses. Even in the intented core development of tech,  the estimated cost for chatbot with active monthly 10,000 users is around 12,000 USD per year. 

Thus, we believe that the revenue we can make from grants and affiliate marketing will be able to cover the total cost as well as they can enhance the scalability of our service. Moreover, we are willing to fund this by ourselves for the first two years. The more we have users, the more revenue we make. In such way, we could make the financial sustainability.