InterviewAI
WEB · AI · HR TECH
Redesign of AI-driven question generator to increase user retention and conversion
Timeline
4 + 1 weeks · 100 Hours
Team
Andrew Chen (me) · Maria Tamayo · Joo Kang
My Role
UX/UI Designer · UX Researcher · Project Manager
Tools
Figma · Figjam · Google Meet · Jira
Context
This product standardizes hiring procedures to reduce subjectively in key hiring decisions.
Constraints
Broad target audience and only 1 developer.
Task Highlights
Collaborating with developers and design pivot due to technical constraints.
ABOUT THE PROJECT
OVERVIEW
RESULTS
What interviewers said about the redesign
“A little confusing at first...But I do like the overall concept that they generate a list of questions that's tailored to my organization's needs and an evaluation score breakdown.”
LZ - Usability Tester
“Clear analysis, clear results...I can see this being a huge dev lift...I like it in that it feels more detailed...providing me with some insight and some visuals that ChatGPT can't.”
KR - Usability Tester
THE OVERARCHING PROCESS
UX IN 4D
IDENTIFYING THE RIGHT PROBLEM
DISCOVERY
BUSINESS CONCERN
Analytics show the product has low conversion and user retention
To inform which research method to conduct, I list key assumptions with the necessary data to validate them.
Conversion obstacles (assumptions)
based on existing user research + stakeholder assumptions + UX team’s audit + 2 competitor analysis
USER INTERVIEWS + USABILITY TESTS
5 usability tests reveal potential causes for low conversion and user retention
I chose a usability test because:
Actual user input (unlike competitor analysis or audit)
Can ask participants for further clarification (unlike existing user research)
Directly concerns InterviewAI's product (unlike competitor analysis or usability testing a competitor)
Testing for desirability and clarity
usability testing with one of five total users
Primary focus: learning interviewer's perceived benefits and thoughts on the signup process.
Users do not gain any significant benefits for using this product
responsible for addressing 2 out of 5 key themes (highlighted below)
UNDERSTANDING THE CONTEXT
DEFINE
MATRIX
Identifying and prioritizing design opportunities
I chose to do a matrix because it helps incorporate feasibility considerations to more clearly define the design opportunity.
Development time = consulted our team member’s 4 year experience as a software engineer (InterviewAI’s developer was unable for this meeting)
Impact = mix of frequency of user comments + consequences for not addressing comments
Assessing our priorities
collaborating with a developer
IDEATING SOLUTIONS
DESIGN
USER FLOW + LOFI WIREFRAMES
Designing for visual focus and meeting user expectations for AI-based tool expectations
Ensured shared understanding of the redesigned product’s function by collectively creating a user flow. To meet deadlines early with extra time to address any unexpected problems, I referenced competitor’s product to design lo-fi wireframes faster.
Understanding the overall flow as a team
the highlighted areas are my responsibilities during this design phase
Addressing key user concern #1
Hard to identify crucial information
Addressing key user concern #2
Limited AI-generated question customization
PIVOT
Pivoted due to implementation constraints despite designs being well-received
Despite assessing technical feasibility using the matrix, the stakeholder preferred phased implementation over a complete overhaul. This prompted us to pivot our designs and focus on simplifying them for easier and more practical implementation.
HIFI WIREFRAMES
Revising the user flow and designing for smoother incremental implementation
We revamped the user flow to accommodate the simplified design. Once the revised user flow was finalized, I iterated on the design, specifically incorporating a scoring system and incorporating more options to customize the AI-generated questions.
Understanding the simplified design's functionality
revising the user flow post-pivot
the highlighted areas are my responsibilities during this design phase
Incrementally refining the designs
focusing on input forms and scoring system
GAINING USER FEEDBACK
DELIVER
USABILITY TEST
Testing reveals major user concerns — I iterated the designs before handing off to stakeholders
I conducted two of the five usability tests, primarily assessing:
Clarity and usefulness of the scoring evaluation for interviewers (a feature the stakeholder plans to implement)
Impact of the current signup process on interviewer drop-off rate (related to conversion concerns)
Interviewer perception of the product's impact on the hiring process (related to conversion and user retention concerns)
Addressing key user concern #1
Product gives mixed signals about intended audience
Addressing key user concern #2
Excessive manual input
CLOSING COMMENTS
TAKEAWAYS
NEXT STEPS
Short-term recommendations for InterviewAI
Primarily a mix of business-related logistics/strategy and incrementally testing improvements to the product.
Immediate next steps for InterviewAI
based on most recent 5 usability tests
OPPORTUNITES
Long-term opportunities for the product
Mostly implementing high development cost features and utilizing new technology to its fullest.
Long-term opportunities for the product
based on most recent usability tests + global trends + existing interviewer workflow
LEARNINGS
Key lessons I learned after completing the project
Working in a team environment provided a different perspective on the purpose of deliverables. It was challenging yet rewarding to leverage deliverables to foster a shared understanding among team members.
