Conversational AI for Patient Intake

Conversational AI for Patient Intake



The Problem

An innovative healthcare technology company in the chronic illness space wanted to reimagine their form-based approach for capturing patient health information by:

  • Increasing the quality and value of information captured,

  • Engaging patients in the process, and

  • Providing a unique experience within their market.

Our Approach

Synaptiq analyzed their existing forms and applied data science and machine learning techniques to historical patient information.  From this analysis, numerous user experience improvements and data clean-up tasks were identified and interesting patterns were discovered in their data.

The Solution

From these findings, Synaptiq designed a prototype application developed for testing the user experience.  It was a novel mobile-first, chatbot experience based on a data-driven algorithm that learns about the patient and asks questions based on previous responses.

Key aspects of the design process comprised:

  • Restructuring the question flow,

  • Rewriting form questions into a more conversational, educational style,

  • Devising a user experience and set of question widgets that fit into a conversational flow, and

  • Developing a command line version of the algorithm for experimentation and early user testing.

From the design, Synaptiq developed the prototype -- an API-driven full stack application using React.js, Ruby on Rails, and Python. The application used the core elements of the design and data algorithm on a subset of the intake forms.

The Results

The team performed two rounds of user testing, made improvements to the application, and closed out the project with a roadmap to fulfill the overall vision.