automated health assistant
personal health assistant chatbot
The real challenge that XUND manages to solve is to hide the complexity of creating a precise diagnosis behind a really simple user flow. Thanks to that simple flow, anyone can get a diagnosis and a clear picture of the causes and illnesses in a few minutes. It can give you a helping hand to decide, whether you should visit a doctor or simply stay at home.
We implemented the following components of this complex system:
The implementation of the chat application and user flow was the most critical part of the solution as we had to implement a fully customizable chat solution from scratch that can handle numerous paths until the user gets the final diagnosis.
The admin application is responsible for managing and customizing the chatbots to suit the needs of different partners. It is able to create new bots and customize existing ones with ease.
The backend application serves as a middleware between the chat and the medical application and acts as a security and data layer for both the admin and chat applications.
In top of that, we created the CI/CD pipeline to have a fully automated workflow.
We created the chat and admin frontend applications in React. We used the latest React features and Redux to build an application that is easily maintainable.
We had to solve issues like having hundreds of different paths, presenting numerous symptoms, illnesses, questions and being able to edit any of the previous steps, jump to any step in the user flow and undo all the changes we made.
As we knew the complexity of the user flow and identified the state management as the main issue from the beginning, we decided to create UI and logical building blocks that can be easily attached into the flow or swapped by another building block. This way we managed to keep the complexity as low as possible and have a readable, maintainable code base that could be handed over in a few days.
As for the UI part, nearly all the components and styles are fully customizable from the admin page. The fonts, icons, images, styles are all dynamic.
The backend application, developed in Node.js, acts as a security, data and middleware layer between the chat and the medical API. The backend application stores all the chatbot configuration settings, provides a REST API for the chatbot frontend and forwards the requests to the medical API.
The XUND and SnapSoft teams worked actively together during the whole project, so the handover went smoothly. Now the solutions is ready to help as many people as possible.
backend and chatbot UI
We proudly present the results of the fruitful association and partnership with XUND Solutions GmbH.