Dialog Manager for Interventions
As part of the DIDEM project of the IMTT Institute, an application for the use of intervention chatbots in healthcare was implemented. For this I designed and realized four key services.
Management of chat dialogs.
This service acts as an interface between the system and the health scientist. Experts can use it to add and configure chatbot dialogs.
Transforming the dialogs.
To enable the system to read the added dialogs, I implemented a transformer as a microservice, which bundles the dialog strings into a JSON format. The information is then persisted in a NoSQL database.
The core component of the application.
This service is designed to retrieve targeted interventions from the NoSQL database and run them, according to the inputs of the intervention participants.
Mobile application for intervention participants.
The mobile application (iOS/Andriod) acts as an interface between the participant and the intervention. It accepts user input, forwards it to the dialog manager, and then receives the appropriate dialog thread from the manager.