The accuracy to which Brobot responds to questions is good, but not good enough to be deployed as a professional service just yet. There are still inaccuracies that can only be overcome by continued testing and training.
It has been found that given the short time frame from data phase to production phase, the bot is performing well. It, however, requires to be fed more data and needs to be able to answer a wider range of questions including but not limited to the costs associated with mental health services. Lack of credible data for such aspects has led to the inability to answer these questions.
This project has vast possibilities of scaling and expansion ahead. The existent chatbot app can be linked with the dataset of all mental health service providers in the specified area and people can be linked to the one of their choice.
With proper and credible data, more options can be added to perform a proper screening (as prescribed by a professional) for users who wish to get a report and send it to their mental health professional. It can potentially fill the gaps along the entire mental healthcare journey, from symptom monitoring to episode management.
