Project Background

What: Reet is a company that is hiring for a marketing manager position. They would like their hiring managers to feel they’ve selected a candidate without the influence of their unconscious biases as far into the hiring process as possible.

During this round of hiring, they've decided to use Luna Connect®, a blind interviewing platform using avatars and Q AI genderless voice technology. The blind video platform emulates human interaction but does not convey gender, age, race, disability status, or sexual orientation. I designed the blind video concept.

Reet's Business Objective: Reet understands diversity increases business development, but the company also wants Reet's hiring managers to choose candidates without the influence of unconscious biases further into the hiring process. 

User Problem: Hiring managers want to feel they have hired candidates without unconscious bias

Solution: Replace a standard phone interview with Luna Connect's blind interview  

Assignment:  Create a minimum viable product (MVP)

Tools Used: Figma, Whimsical,

Areas of Highlight in this Case Study 




Why Blind Interviews?

I have a lot to say on this matter, so please read more on the background of this case study through my Medium article: 

In short: through the process of redacted resumes, blind chat/emailing, an initial blind interview- the hiring manager will feel confident they have selected the final candidates with the least amount of bias as possible.


Market Research

What's out there already that uses AI? 


AI objectively assesses soft skills and personality traits through blind interviews with Tengai. Weeds out candidates based on the customized criteria set for the role; AI scoring.


“Every candidate gets an interview.”

Candidates text on mobile in their own time with conversational chat AI. Every candidate gets insights back after the interview.


AI analyzes recorded interviews that candidates answer via pre-recorded questions. There is never a face to face real-person interview.

Caveats with these programs:

-Robot interviews don’t save time because recruiters/hiring managers go back to watch the video anyways

-A person can redo their response before they actually submit- this doesn’t necessarily demonstrate a person's true communication ability (but does it since we can edit after we send?) 

-It's impossible to create an unbiased AI since humans design the system


How About We See What Problems People That Hire and Apply for Jobs Face?  

Interviews and Surveys

A major constraint on this project was that I did not interview the “right” people at the right times. I initially started out surveying with general questions on interviewing, as a hiring manager and applicant, to see what key takeaways there were from these surveys  

I then interviewed a handful of people from my boot camp.

Of the interviews, someone emphasized their desire that their company focuses on DEI.

When asked if they would want a blind interview most said no, they would want the real 1:1 interview because soft skills are just as important as hard skills.

Most were dubious on the blind interviewing technology; how well it works and how much soft skill can come across via avatar?

Quote From Interview Participant

"I would rather know as soon as possible into the interview process if there would be discrimination. I want to be SEEN right away."

Key Takeaway

You can't ask the user what they want because they don't know. 

 I waffled quite a bit on whether to continue with my hypothesis that the user would want a blind interview. From the people I interviewed in my boot camp, most had said they would not want a blind interview. I tried to source hiring managers/recruiters to interview to understand what they would like to see improved in the interview process to make it more equitable. Despite posting within various Slack channels, Facebook groups, LinkedIn, and reaching out to peers, I was not able to obtain direct interviews with hiring managers in direct relation to more equitable hiring practices. 

I posted on Reddit whether I should design for an unwanted idea and gained a lot of valuable feedback from the Reddit community

So, I continued on with my hypothesis. 


The Idea 

A Typical Hiring Process:

-Submit resume and cover letter
-Phone interview
-In-person/live video interview with hiring managers


Reet’s Blind Hiring Process Using Luna Connect®:

-Redacted resumes and cover letters 
-Blind chat AI communication and emailing platform
-Blind video interview with a hiring manager (the focus of this project)
-Final in-person interview with hiring managers 

Reasons Why This is a Good Approach

This would be a better initial interview process than robot interviewing, as it’s more personal. 

The hiring manager and applicant still get a sense of each other's character while mitigating unconscious biases, compared to a phone interview where you can't see body language and can be influenced by unconscious biases. 


User Persona

Hiring Manager X

X, a hiring manager for Reet, has decided to use Luna Connect®, a blind hiring platform to hire their next candidate for Marketing Manager. 

- They would like to build a diverse team but don't want the pressure to check off diversity boxes to influence whom they bring in

-X wants to utilize a platform that helps mitigate the influence of unconscious bias as far into the interview process as possible


I wanted to test to see people’s thoughts once having seen the idea for the blind interview.

The test wasn’t as much to see how well built the app could be; it was more testing the concept. 


-Mid-fidelity wireframes

-Blind interview in lieu of phone interview; which would then follow with in-person final interview

Key Results from Testing

I gained surprising feedback on this type of blind interviewing via testing. 

Question to participants:

You have gone through the demo of the blind interview process.

Would you choose a standard phone interview or a blind video interview?

Out of 15 people: 

8 would prefer a standard phone interview

7 would prefer a blind video interview 


Motivations on their choice between standard phone interview and blind interview:


"I would choose a standard phone interview, so I could leverage my character + tone to show interviewers what kind of personality I have. If it was a large well-known company I would still apply, but nothing less. I feel like this could make me feel like I was being scammed without talking to /seeing real people."

"It's a cool option to have an interview. I think I would prefer an initial standard phone interview because at least we can tell the personality of the potential coworkers you will be working with. If a company approaches this way of interviewing, I would still apply for the job because if the job is the position I like, then I would try it out.


"I like the process. I would definitely apply with blind interview, because as an introverted person I don't want people to have a judgemental side about my appearance or anything else except my knowledge and experience."

"Q AI is super cool to implement. It’s creating the equality in recruiting."

"I think I would like to try the blind video if there's an option. Cool idea. I think this might be more unbiased."

“If I was a hiring manager, I think it could be helpful with an unbiased decision.”

“It's an interesting concept because it's different and it promotes an equality.”


"I would be hesitant.  I think that there are too many ways that AI can screw people over, such as women's voices not activating voice recognition software.  And my doubts increase the more layers of tech there are.  So in this case, there's the voice editing, but also the expectation that the animations will mimic one's face thanks."

“I think there are small things like facial expressions and intonation in voice that would be missed in the masked video interviews. So I'm not sure.”


From my initial interviews, people said they would not like blind interviews. However, after testing, it appears with a more trusted interface and technology, candidates would be keener on this idea.



Mid-fidelity wireframes


I overlooked adding the Help Center in the navigation bar during the testing process, so added this in afterward. 



Updated the avatars to not look quite as robot-like. I designed the little avatar bodies myself.


Avatars need to not demonstrate: age, sex, gender, race, sexual orientation, disability status, location, and accent. 


As mentioned before, I did not interview the “right” people at the right times. 

Not proper demonstration of the technology I am proposing to utilize for the interview process. Ideally I would have had a simulated interview with the avatars speaking through, Q AI, the genderless voice technology.

Not possible to have completely unbiased influence in an interview. Even the way people talk and write is indicative of their background, whether if English is their first language or how well trained they are in writing.

Post Mortem

One of the most valuable comments from the Reddit post that I'll always carry with me as a designer is: If you had asked people years ago whether they would have wanted Netflix, they would have said, "No way, I like going to Blockbuster." 

I had learned this through my studies so far on the UX research process but really saw this fact come into play. Also, considering I was not interviewing and testing with my ideal target audience (hiring managers/recruiters), I was surprised to see that almost half of the testers would have chosen the blind video interview over the standard phone interview, after going through the prototype testing. I think more people are coming to realize how much influence bias has in our everyday interactions and choices. 

To the concern from some interviewees/testers that this process seems impersonal:

Again, what is "human interaction" these days as we know it?