AI resume tailoring
Helping job seekers write tailored work descriptions to specific jobs
Overview
I spearheaded designs to help job seekers seamlessly tailor their resumes to job descriptions. This project won two awards, “People’s choice” and “Best use of design system” during the 2023 Indeed hackathon. After winning the hackathon, our team pitched this project to senior leadership and championed its integration into our product roadmap, ultimately shipping on Indeed.
My Role
Product designer
Team
Software engineers
Data scientist
Product designer
Platforms
Desktop, Mobile web
Timeframe
3 Months
Impact
Problem space
Knowing how to craft and explain your work experience into only several resume bullets is a daunting task. From research with Job seekers, lack of knowledge about what constitutes a strong resume is a major barrier to creating high-quality applications.
"I'm unsure how to efficiently present myself with words. I also don't have enough to put down for a resume."
Tharikaa, Job seeker
Refined problems to solve
Problem 1
Manual process of tailoring a resume
Starting from a blank canvas is a daunting task when writing a resume. Without guidance, job seekers are on their own to decipher how to wordsmith their work experience.
Problem 2
Lacking confidence in resume content
Job seekers are unsure how to best tailor their resume work experience to job listings. Even those with pre-existing content found difficulty in framing.
Problem 3
Inconsistent existing suggestions experience
Suggestions varied wildly across product surfaces, creating a fragmented experience for job seekers. The opportunity was to unify this guidance into a cohesive system.
Opportunity
How might we help job seekers best represent their experience to show employers they're qualified for a position?
Aligning internal experts around a unified vision
To ensure early alignment with cross-functional experts in our space, I led a workshop between UX, engineering, and product teams to begin our discovery phase. Showing our audited end-to-end experience brought eye-opening insights and identified major knowledge gaps to better define as a team.
Resume suggestions workshop
Outcomes from the workshop
I worked alongside my content design partner to turn the workshop insights into actionable items for our product roadmap.
Outcome 1
Defining criteria for quality resumes and suggestions
Understanding what defines a quality candidate was key to utilize as a baseline so that we could recommend what improvements a job seeker could make to their resume.
Outcome 2
Simplifying the actions for job seekers to improve the content of their resume
Given the years of data Indeed has on resumes that receive positive outcomes from employers, we can utilize a data-informed approach for how to best craft a resume.
Outcome 3
Utilizing occupational segments to make more relevant recommendations
We surfaced resume suggestions based on job seekers past experience However, research showed that career changers received irrelevant suggestions when starting a new search.
Shipping an MVP with limited team bandwidth
After the workshop, I began triaging top opportunities with my team and identifying high-impact initiatives. We started with high impact, low-hanging fruit opportunities to scope out engineering work for the quarter. These were promising experiments which improved discoverability of suggestions and drastically increased resume edits by +13%, showing an early clue that light guidance led to action by the job seeker.
MVP hypothesis
Improving discoverability of existing suggestions will help job seekers fix mistakes on their resume.
Outcome
Crafting a design vision with limited scope
Our team had ambitious ideas for improving resume guidance, but resource constraints limited us to short-term fixes. To work around this, I used our company hackathon to prototype a bold vision: AI-powered resume editing that actively fixed wording issues, rather than passively explaining them. Combining past user research with Indeed's job description data, I built a compelling demo that earned buy-in for a larger initiative.
Hackathon demo
Job seekers can leverage their existing work experience bullets to be refined and tailored automatically to the job description, drastically increasing the resume keywords matching to a job.
Outcome
Long-term milestones
Visibility and acknowledgement from winning the hackathon helped build a strong case to build out this vision. I then pitched and worked with my senior product leadership to influence milestones for a beta launch. Moving from hackathon proof-of-concept to live MVP, our main hypothesis to test was if AI tailored work experiences were of perceived value to job seekers and positive outcomes improved.
Resume.com implementation
AI tailored resume feature based upon our team's hackathon project
Job application and Indeed integration
In-context guidance while applying to jobs on Indeed
What I learned from this project
This project was exciting for me as a designer, since I found ways to creatively go beyond existing implementation and roadmap constraints to influence bringing a bold vision to life.
Don’t feel constrained to the existing roadmap and limited bandwidth. Through prototyping, design can show the vision and be a change maker to bring excitement to the team.
Creating a diverse skill-set team not only makes a stronger product, but also ensures various angles of a problem are considered and addressed.