Helping job seekers prep for interviews.

8 Weeks
UX/UI Design, UX Research

Preparing for upcoming job interviews is difficult.

An overwhelming 93% of Americans have experienced anxiety due to job interviews, according to this one article by JDP. 41% out of the 2018 surveyed participants reported being most nervous about not being able to answer a difficult question.

"How Might We" make it easier for job seekers to ace their next interview?


Help people prepare great, personalized answers for potential interview questions.


But first, what interview resources are already out there?


Consists of practice sets and video modules for interview preparation.


Provides anonymous reviews and resources about a company, including questions asked in that company's interviews.

Manual Research

Gathering resources from Google, company website, LinkedIn, etc.

*Recently, there has also been a rise with everything "AI," including interviewing tools made with AI. Although they may not be as widely used at the moment, they are likely to become more popular as go-to resources very soon.


Despite there already being existing products and resources, many job seekers are still struggling with interview prep.

Method: Survey, affinity map

I wanted to dig deeper into what specifically job seekers were having trouble with when it came to preparing for upcoming job interviews. I conducted a brief survey with 6 participants, asking about their experiences/challenges when preparing for interviews.

Affinity Map

Major Insights

  • It's difficult trying to prepare for unknown interview questions.
  • It's difficult trying to word "great answers" to these interview questions and not generic ones that the interviewers have heard before.

How might I make a better solution in the midst of these existing products/resources?


Incorporating AI + Glassdoor

Glassdoor has a unique pool of resources for job seekers, including a database of interview questions shared from people's personal interview experiences with specific companies as well as various collections of honest reviews about companies and their cultures, provided by anonymous past and current employees. With AI, Glassdoor's valuable data will be compiled into a simpler, straightforward guide to preparing for interviews.

Instead of creating the tool as a standalone product, I decided it would be a great integration into Glassdoor's platform as a new feature.

Introducing the Main Feature: Answer Generator

Using Glassdoor's database of interview questions, the AI-powered Answer Generator will generate a list of potential interview questions based on the company, job role, and customization preferences. Users can then choose to generate sample answers, which will be more personalized according to the user's resume. In addition to the sample answer, the generator will use data from past and present employees' reviews about the company, as well as company's goals and mission statements, to identify what the company might value in a prospective candidate and how the user can emphasize those values in their interview questions. This will come up alongside the sample answer as "Tips."

Desktop screen mockup of a conceptual feature for Glassdoor called Interview AI.

Where does this fit into the job interview process?

Once a job seeker lands their first interview offer from a company, they can start using Interview AI in order to see what questions to expect and how they can answer those questions. The personalized answers from the Answer Generator will be a starting point for how they can word their interview responses, since personal anecdotes cannot be made up. Along with the sample answers, the generator will provide any useful tips and key points that can be used in their interview responses in order to make the interviewee seem like a better fit for the company.



Onboarding for Different Users

I have identified 2 types of users: people who have a filled out specific preferences in their Glassdoor profile, and people who are generally new. To better accommodate the different users, I created slightly different onboarding experiences.

Refining the Answer Generator

I reduced the amount of information that a user has to fill in when using the answer generator by having the user pre-fill more of their preferences from the onboarding experience. (These preferences can be updated in the profile section). I've also changed the user flow so that the answer generator allows the user to generate answers for specific questions and also gives the option to re-generate the answer if they aren't satisfied with the initial response.


In the low-fidelity prototype, "My Answers" would show all saved answers without way to filter them. I've added a company name dropdown so that saved answers will show depending on the company that the user looked up previously.


The Final Design


For those who have already filled out their preferences in their Glassdoor profile, these preferences will preface during the onboarding experience for confirmation and are available to edit. For those who haven't, they will be prompted to fill in their desired job title and desired location.

AI Answer Generator

The user can switch between companies for company-specific questions. Once they fill out their Q&A preferences, they can hit "Generate Questions" and a list of possible interview questions for that company will pop up. Clicking "Generate" next to the question the user chooses, the AI answer generator will personalize a sample answer for the user according to their resume as well as any key points that the candidate can include in their answers in order to present themselves as a great fit for the company. The user can re-generate the answer to their liking and hit "Save answer" once they are satisfied with the result.

*These personalized sample answers serve more as guidelines/starting points on how to answer interview questions according to the user's past roles and responsibilities. The user should bring in specific details and examples from their own experiences to tell their authentic stories.

My Answers

All the previously saved answers will show up under "My Answers" tab. The user will be able to filter it out by the company they searched questions for. Each answer is editable so users can be more specific and bring in snippets of personal anecdotes to refine their answer.

What I learned:

  • Sometimes there are already existing solutions to problems, but there will always be room for improvement
  • Don't be afraid of iterating and making changes, even if it means having to get rid some things

Moving forward: 

  • AI's impact on data privacy is an ongoing, complex conversation, but new regulations related to privacy and AI are imminent. As new policies are developed with time, I believe it's important to consider the application of new AI policies as well as incorporating privacy principles into the product development.