Zoe | Devpost

Inspiration

VIDAVO operates in the field of ehealth / mhealth for fifteen years, linking citizens and healthcare providers over its intelligent monitoring and care management platform Vida24®. VidaMe is the “wellness” cornerstone of Vida24 for empowered consumers, enabling behavourial change to healthy lifestyle, with the use of a smart app, connected devices, and a cloud platform, where big data & predictive analytics utilized to provide coaching and motivation. The Zoe bot, could provide advices, information and automated nutrition plan depending on metabolic profile of the consumer, guidelines and predictive analytics for a healthy life. The main differential point is that Zoe Bot, provides automated personalized nutrition plan and recipes, depending on consumer needs and health indicators. The consumer has the ability to use the Zoe standalone via Facebook messenger but can combine the VidaMe app for tracking health and general wellbeing related parameters, via structured questionnaires based on goal setting theory, regarding nutrition, physical activity, vital signs and other health indicators. Our target is to combine Zoe with VidaMe app, as a digital assistant and guide to use the app and as a nutrition/health advisor.

What it does

Helps people to set their lifestyle goals depending on metabolic profile and improve their health outcomes.
Zoe gets useful information from the consumers like gender, age, height, weight and build metabolic profiles; provides information messages and advice to consumers; creates automated nutrition plan depending on metabolic status, calories and personalized nutrition goals; measures the daily progress;

How I built it

The Zoe bot is deployed in Facebook Messenger and instance response in a user-friendly format card showing information to the consumers.Zoe bot developed using Java 8 as the native code language. The intents and slots created by using Amazon Lex tools and built an Amazon Lambda function in order to create the business logic of the bot. Slots values validated in lambda function, and send the response to Lex platform. The data retrieved from an Amazon Dynamo DB database using userID as a unique key to identify the user each time he/she communicates with the bot.

Challenges I ran into

VIDAVO’S IT team have experience in development of mhealth apps, platforms, api’s and services. Building a nutrition bot combining new technologies and health knowledge it's very challenging it. It was our first approach to build a bot and we didn’t use before the Amazon Lex and Lambda, so our main challenge was to learn very quickly to manage all the components.

Accomplishments that I'm proud of

This was our first time we developed a chatbot. Winning or not, this was for us a good opportunity to investigate new technologies and tools to a new field in real time. The combination of data complexity and the sensitivity of health related data processed of a digital assistant in the healthcare sector is a big challenge. We are very proud, that we did our first step and ancompishment this challenge using AWS services.

What I learned

Getting familiar with Amazon services and tools such as Lex, Lambda and DynamoDB, we believe that is the most beneficial thing we learned during this challenge. We gain knowledge in a field that we are very interested in, the chatbots industry. Getting familiar with future tools that we can use to change the way our customers interact with our apps is very important to us.

What's next for Ζoe

Healthcare chatbots will spread around in many areas of healthcare, digital health assistants in the form of chatbots are already able to answer basic questions about health management. Our scope is to integrate Zoe, to our VidaMe app and platform, as an assistant via chat and voice. Also, we want to combine knowledge, connected to 3rd party API’s ex healthvault, food product database, Databases ex CDC, WHO, NHS, and NIH databases in order to provide more accurate data and motivate the consumers to live healthier.

We will proceed the development of Zoe bot in order to provide:

automated plan not only in healthy people but also in diabetics, CVD Diseases, Obese people who are using our platform;
advices to the users and motivational messages in order to help them to reach their nutrition goals;
useful information about their nutrition plans and the usage of the platform;

The Zoe bot could be our intelligent assistant for worried-well citizens and patients.