Interview Engine at Ethos
Ashwin Pattan, Software Engineering Manager
The interview is the first step in an insurance application. It is the process of collecting all the required data from the customer for us to underwrite the application. The interview not only serves as a digital replacement for a traditional paper application, but also enables us to ask the customer the right follow-up questions in real time and ultimately provide an instant underwriting decision.
Similar to a traditional application, there are different sets of questions that need to be answered based on the product that the customer is applying for. The set of questions are regulated by the insurance carriers, and any updates or modifications need to be approved by the corresponding governing bodies.
The user interview at Ethos is custom built to support our automation goals. The interview is designed to collect granular information that can be fed to the underwriting engine to produce accurate underwriting decisions.
Interview Engine at Ethos
An interview engine is the collection of services that helps Ethos product and actuarial teams create and serve interviews to customers. The interview engine and its configuration is of paramount significance in helping Ethos achieve the goal of fully automated underwriting.
The primary role of the Ethos Interview engine is to serve the questions and answer choices (where applicable) to the applicants. The engine provides an optimal experience to the applicant by 1) asking the required questions that help determine the applicant’s eligibility for an insurance product and 2) asking these questions in real-time to eliminate the weeks of back and forth in the traditional application process.
The key responsibilities of the service are:
- Product-specific, regulation-compliant interview questions
- Seamless interview continuation for returning customers
- User input data validation
- Answer driven follow-up questions
- Collecting user answers for all questions
Features
Here are some of the key features of the interview engine.
Selective Questions
The interview engine determines whether a question needs to be asked or not, depending upon customer profile and/or answers. A good example of this is that any pregnancy related question/answer choice is not relevant for a male applicant.
The interview engine filters out these irrelevant questions/answer choices when deciding what the next question (and associated answer choices) should be.
Follow-up Questions
The interview engine decides based on the applicant’s answer what the next question should be. Instead of asking a set of questions in pre-defined order, the engine contextually asks follow-up questions based on the answers.
For example, the next question to an applicant’s selection of cancer would be to collect additional relevant details. Conversely, an applicant who says they do not have cancer would be sent on to another topic.
This dynamic nature of the interview makes the applicant experience contextual and ensures we collect as much detailed information as possible to facilitate automated decision making.
Seamless Interview Continuation
User interviews can sometimes be long, based on the path taken by the applicant. Sometimes, the applicant does not complete the interview in a single session, especially if the interview involves follow-up questions. Such applicants tend to return to Ethos to complete the application, and we ensure that they do not have to redo the interview. Every applicant who dropped midway through the interview can return to pick up the application right where they left off.
Versioned Interviews
An applicant’s journey through a version of the interview would not change, even if a new version of interview is released for the same product. The version of interview is determined at the start of an application, and is retained for the application lifecycle, i.e., application to underwriting to policy issue/close.
This is a very important feature for compliance. Applicants could be in different stages of the interview, and if we switch to a new interview version, there’s a possibility that we wouldn’t ask the applicant the new/modified question. That would lead to compliance issues, where underwriting uses partial/incomplete data.
Architecture
The core logic flow of the interview engine can be summarized this way.
Configuration System
The interview engine at Ethos also comprises a configuration management system. It’s used to generate interview configuration, and decides the flow of the questions in the interview. After a new question is approved by the governing bodies, the product and actuarial teams can update the interview configuration to create a new interview version for a specific product. The configuration system provides a visual representation of how the interview tree would look like for the applicant and the author can navigate the tree to ensure that the configuration is created as intended.
This interview configuration is then tested in our testing environment, and validated before releasing the version.
The Future
The interview is one of the most common interactions that a customer has with Ethos. The user experience delivered during the interview directly impacts the customer’s decision making process in applying for a product with Ethos.
One evolution path is for the engine to identify the risk profile of the applicant and determine the interview configuration. The follow-up questions could help clarify the risk. Understanding the risk will help navigate the applicant towards an approved application. This would need a real-time active integration with the underwriting engine. That way, we can determine what factors counterbalance the risk profile of the applicant.
In the end, it’s all about asking the right questions!
Ashwin Pattan joined Ethos in October, 2019 as a Software Engineer, and is currently the Engineering Manager for the Underwriting platform. When Ashwin isn’t coding, managing or setting goals at Ethos, he enjoys playing futsal, traveling, cooking and watching movies. Interested in joining Ashwin’s team? Learn more about our career opportunities here.
