Details:
Team: Sanket Shukl, Daniel Thomson, Arjav Badjatiya
Responsibilities: Sketching, Ideating, Redesigning Wireframes, Designing Pitch Deck
Duration: 1 week
Type: Graduate course project under Prof. Marty Siegel
Prompt by InnovateMap:
Our CEO keeps saying it's "Uber for Doctors" but we have no idea what that means. He says that he wants to transform health care in the Midwest. Moreover, the CEO wants us to design a chat bot for this!
We were working for a healthcare startup looking to design a service to help doctors connect with patients. Also this comes at a good time technologically. Bots and conversational UI's (ConUI) are starting to come to life. Slack, Facebook Messenger, and the new iOS 10 have been massively transforming the messaging/chat medium and there are endless possibilities. So the deliverable for us was a high fidelity ConUI design of one primary use case.
Research Insights:
- Also, there exists numerous services that are trying to make the doctors available to the patients quicker and faster. How can a new startup design a disruptive technology leveraging the latest chatbot trends to innovate and make the experience even better for the patients?
- There are way fewer doctors per patient in the midwest as compared to the other parts of the country. In spite of this, how can we make the doctors and the pharmacist more connected to the patients?
What we focused on:
To break the problem down, we focused on the end part of the chain on interactions that a patient goes through in the entire experience of visiting a doctor. If we could innovate on making the interactions between the pharmacist and the patient more friendly and direct, we can make the entire experience better.
Also, our research showed that patients have a lot of questions at the end of the visit to their doctor, and the pharmacist can act as a facilitator, instigating the patient to get better quicker by prompting them to do the right thing.
We plan to achieve this with the help of automated chat bots and applied NLP.
Why we designed a chat bot for pharmacists:
“Pharmacists are the bridge between what a doctor wants and what actually happens . . . we have a huge responsibility to ensure patients are properly informed about their treatment and that they are doing what they are supposed to.”
● Doctors have a very busy schedule. The time of the doctor is much better spent attending the expert needs rather than other admin tasks.
● In the midwest, the doctor to patient ratio is low. There are way less doctors per patient as compared to other parts of the country.
● The tasks of the doctor and not worth replacement with any kind of advanced AI or applied NLP.
● The ultimate goal is to make it easier for the patient to go to the doctor for the next time, in turn making the patient feel more cared for.
● This would increase the satisfaction level of the patients by making them feel connected to their ‘care team’.
Design Walkthrough:
After his appointment with the doctor, Jacob picks up his medications from the pharmacy and is then asked to sign up for the service to help facilitate his recovery.
He receives a text message from the service, introducing Dawn- his personal digital assistant.
The chat bot lets Jacob select from a predefined set of replies or type his own. If Jacob wants to use the text prediction, he can quickly chat with the assistant without having to type.
The assistant gives Jacob some starting options to engage him in conversing with the bot.
Jacob can simply respond by clicking on one of the options or by typing in the option number.
The chatbot aims at bringing the UI to Jacob's fingertips. When Jacob asks to set a reminder, the chatbot asks for the time in two ways- either by interacting with the UI pulled by the chatbot from the reminders app or by telling the chatbot the time and date.
Once the reminder is set, a feedback will be given, with an option to revert the action or to edit the reminder.
The next morning, Jacob receives a text as a reminder he had set yesterday in addition to the native reminder app reminding him.
The chatbot also sends time and context aware message, like texting Jacob on the weekend about the reminder to control alcohol consumption.
Day 15:
Later during his medication period, Jacob notices a rash on his shoulders. Since the chatbot is his personal point of contact with the pharmacist, he can send a message, to which he receives an automated response to gather further information about his condition.
Once Jacob responds with more information, the chatbot will respond accordingly.
SInce Jacob's condition is recent, the next information he needs is a visual input of the rash which is going to help the actual pharmacist to recommend medication updates.
The chatbot clearly states what part of the conversation is a human and what part is automated, to build trust with the user.
Once the image is received, the chatbot passes over the control to the pharmacist with all the information delivered to the pharmacist.
Now the next message Jacob receives is from the pharmacist along with a few updates he needs to make to his medications as clickable cards.
Jacob can decide, without leaving the chatbot, what medication he wants to purchase and a confirmation is shown as feedback. He can pickup the new medicine from the pharmacist.
At the end of the recovery cycle, the chatbot will send an automated message reminding Jacob about his follow up meeting with his doctor, providing him the date and time of his scheduled appointment.
Jacob can then choose to reschedule it or let the chatbot know he will be there.
Since Jacob has asked for rescheduling, chatbot gives him next available times for his doctor.
Now that Jacob needs to reschedule his appointment, he can choose from the next available times for his doctor by responding to the chatbot message.
When he selects a suitable time, the chatbot will send him a confirmation of his new appointment.
Also, the appointment will be added to Jacob's calendar, and a confirmation card will let him revert it or edit it.
Personal Reflection:
My reflection coming soon. Please be sure to check back.