MeDiCo Chat

MeDiCo Chat is a Customer Care, Work Collaboration & Marketing Tool Artificial Intelligent Chat Bot for Medical Facilities and Health Consultants.

Comment 1

Inspiration

I once lost someone close to me due to poor medical service delivery as a result of scarce personnel and lack of information. In Africa, there is high shortage of personnel in many private and local hospitals. Research I carried out also shows that majority of these facilities do not have online presence. And where they do, the information available to the public with regards to customer care and service delivery is nothing to write home about.

I have been working on a project at my spare time called Medical Digital Consulting (MeDiCo). One feature of this web and mobile App is what I called MyHealthChat. It's a special effort to promote the use of chatting technology to improve medical consultancy and medical service delivery. The primary objective is to save lives.

But because many medical professionals are usually busy, I figured out that a well design AI chat bot will help to fill the vacuum. While I was still thinking about how to go about designing the messenger bot, I stumble on this hackathon. So, I have to suspend the main project and get on with this as it serve as extra motivation. When all is done, I'll be able to incorporate this into the main project.

What it does

MeDiCo Chat is a Customer Care, Work Collaboration & Marketing Tool Artificial Intelligent Chat Bot for Medical Facilities and Health Consultants.. It is design for medical consultants and service providers to improve customer care delivery and workplace collaboration.

The Chat bot make it possible for prospective and existing clients to get all their inquiry and complains taken care of even without transferring protocol to a life agent. With hard coded personas, the user get the impression that they are chatting with a real agent.

The staff area requires access code. Staff can use the platform to interact with the HR department and submit important report, survey, suggestion, complain or vote for organization polls.

How I built it

I studied the sample bot suggested by Facebook but it doesn't give me the experience I was expecting. Besides, I was looking at building an application that can be in sync with my existing project and can be scaled up later. After some researches, I settled for Node.js and I am using Asynchronous Javascript to make sure that the code is Promise based. I have to design everything from scratch using the same structure as I am using for the main project.

I divided the menu into three parts. The Customer Menu attends to everything about customer care inquiries. I manage to include an appointment booking session, a carousel of services, Billing & Fees information, FAQs, Enquiry, Call Extensions and a pre-loaded medical emergency and first aid section.

The staff area require an access code to use it. Each staff will be assigned an access code. This section is designated for work collaboration. Staff can interact with the HR department, read safety rules, emergency info, submit report, take survey, vote, download files, make suggestion and complain and so on.

The last menu capture neutral visitors who are neither customers or staff. I also hardcoded persona into the project that mimic the change of protocol. The timing of each chat is design to suggest that it is real time typing.

Challenges I ran into

I ran into the challenge of how to use some technology like Async Promises in the project. Deploying to Heroku was also a challenge. But my biggest challenge was using React as the UI for the GUI webview. I also have the challenge of how to handle real-time input. Due to shortage of time, I will have to look into some of these issues in a future version.

Accomplishments that I'm proud of

I have used some existing tools like Chatfuel to create chat bots before but I never knew I can actually design one from scratch. Developing this myself has also enabled me to leave room in the App to scale it up with AI when I am done with the AI course I am currently taken. I can't also forget the 'WOW' look on my friends face when they see the App in action.

What I learned

I learned how to use Node.js and Express.js to build Messenger Bot. I also learn how the GUI webview works and how to use React as UI for the mobile view. This project has opened my eye to how Facebook API works and I can now confidently build Apps across Facebook products.

What's next for MeDiCo Chat

The next version MeDiCo Chat 1.1 will see the introduction of the following features:

  1. Introduction of COVID-19 Module in the Medical Emergencies section for customers.
  2. Adding Facebook page Comment-to-Messenger feature for lead generation.
  3. Introduction of GUI Webview with React UI for the staff area on report, forms, survey, poll and voting.
  4. Introduction of Postgres relational database with Sequelize ORM for improved customer care experience with AI features to enable users utilize some existing API in our project.
  5. Introduction of GUI Webview for the customer care FAQ section.
Share this project:

Updates