Raven AI Job Agent

PRODUCT DESIGN

ROLE | LEAD DESIGNER
CATEGORY | PRODUCT, APP
YEAR | 2025

OVERVIEW

I led product design and product market fit (PMF) for Wellfound’s first AI-powered job search agent, Raven, by driving creative strategy and cross-functional collaboration across product, data and engineering partners.

ROLE/RESPONSIBILITIES

  • Design Strategy
  • Qualitative User Research and Testing
  • Visual direction and design system of product brand and experience

The Context

With the job market's cycle of layoffs and influx of candidates, research showed that candidates struggled to find jobs and spent excessive time searching. Current job platforms and boards were all alike, appeared basic and only scratched the surface for what job seekers needed to make their search better.



While our job search engine at Wellfound provided valuable opportunities, it didn't alert candidates to all relevant roles outside the platform, leading to missed chances. We envisioned a future where job searching was easier, more personalized, powered by AI, and without the grunt work of older systems. After preliminary research, we developed a vision that guided Raven's journey towards our Beta launch and product-market fit (PMF).

The Problem

The job search process is repetitive, labor-intensive, and lacks critical insights, with candidates spending 10-40 hours a week searching.
Current AI tools offer basic, reactive support but fail to understand candidates’ unique skills and goals, leading them to seek external resources for career development. These solutions automate tasks without providing the personalized guidance needed to boost candidates' chances and professional growth.

Key Challenges

Saturate market and timing

  • The chances of getting noticed without a referral decrease if applicants don’t act fast. Applying within 48 hours of a job posting is crucial.

Lack of personalization

  • With a surge of job seekers due to layoffs and competition, candidates face difficulty finding roles that align with their unique skills and goals.

Excessive time searching and applying

  • Candidates spend 10-40 hours a week searching, altering resumes and applying, reducing valuable time for other opportunities.

Research & Sketches

To validate our problem statement and uncover user needs, I led a mixed-methods research phase:

  • User Interviews & Surveys: Conducted in-depth interviews and a user survey to understand pain points in the job search journey—time spent, favorite tools, and moments of frustration.
  • Competitive Analysis: Benchmarked existing AI and job-search platforms to identify opportunities for differentiation.
  • Affinity Mapping: Synthesized qualitative data into core themes and user personas, highlighting needs around real-time alerts, personalized insights, and resume guidance.

Armed with these insights, I sketched low-fidelity workflows for key Raven features—onboarding, alert preferences, match dashboard, and resume editor. I iterated these sketches in rapid cycles with the PM and engineering leads, then translated the refined concepts into clickable Figma wireframes for early usability testing. This research-driven sketching laid the foundation for Raven’s intuitive, need-focused design.

I wanted to set a good design foundation by creating some basic styles, tokens and guidelines for the product including an Identity that was more modern and bold than what was typically seen. More polish was added along the way as we created solutions for our challenges.

Job Alerts

Reduced Time Spent Searching for Relevant Job Opportunities

  • Designed the job agent to scrape the web for relevant opportunities and deliver as soon as they are posted, ensuring candidates never miss out on potential roles.
  • Delivered curated job alert emails on a daily or weekly basis, with key role information and personalized fit, sent directly to users’ inboxes, saving valuable time.
  • Provided company insights to help candidates quickly gather relevant information, reducing the time spent on research.

Email Alerts

Immediate notifications via email alerts allowed users to act as soon as jobs are posted

Company Insights

Company insights made readily available to help reduce research time and give candidates an understanding of the company mission, size, funding and more. Simply tap or hover the company name or logo to view.

Personalized Suggestions

Provided users with personalized job alerts and feedback

  • Created a simple onboarding process to capture essential user information like skills, experience, and interests, which helped train Raven for more accurate job matching.
  • Parsed and analyzed resumes using our very own backend engine, Spider, to show users their fit level and how their skills and preferences align with specific opportunities.
  • Delivered tailored resume tips to help candidates optimize their resumes for better visibility in company ATS and AI systems.

Role Insights

Roles insights shared how well the role matched the users resume and preferences.

Click to try the static prototype

Testing PMF

We used the Sean Ellis score to measure our progress toward product-market fit (PMF), with 40% indicating that we had reached PMF. Our survey results showed that 28-31% of respondents selected "very disappointed," indicating we still had work to do. As a result, we planned to conduct more UXR interviews to gather additional insights.

Based on user feedback, beyond just landing interviews, users believed that if Raven could handle the grunt work—automatically customizing resumes, enabling easy editing, and even auto-applying—it would significantly enhance the experience. For many, it would mean the difference from being somewhat disappointed to very disappointed. In fact, some even expressed willingness to pay for such an experience.

Pricing Experiment & Learnings

To validate users’ willingness to pay for Raven’s “grunt-work” features, we ran a pricing experiment at the end of a free trial:


  • A/B Test Pricing: Offered two subscription tiers—$5.99 and $9.99/month—after users’ trial expired.
  • Optimized Trial Length: Shortening the trial from 14 to 7 days increased conversions, keeping Raven’s value top-of-mind.
  • User Segments:
    Urgent-need users (actively job hunting) generally subscribed immediately.
    Casual users supplemented Raven with other platforms offering free features, indicating demand for a freemium tier.
  • Early Results: Secured 70+ paid subscribers in this small-scale test, confirming market appetite and informing our roadmap for feature enhancements and free+tiered pricing.

Continuing to strengthen Raven’s visual identity, I created a suite of custom graphics, illustrations, and data visualizations—that are seamlessly integrated throughout the product experience. These assets reinforce Raven’s brand personality and guide users through each feature with clarity and consistency.

With feedback from over 1,600 beta users and insights from our research study, we were confident we were on the right track, evolving the product toward PMF and its post-beta release.

Honestly, Raven is so much better than the lists of job-openings I get from other recruiting platforms. Every job Raven suggested was interesting to me. Fantastic use-case for Generative AI, and well-implemented. Great work.

USER QUOTE

I have really enjoyed using Raven. I feel like it does a generally good job at summarizing and comparing the job description to my current skills and experience. Probably my favorite tool for job searching right now. Feels passive and like it weeds through some of the jobs for you.

USER QUOTE

Raven has given me two jobs to apply to and they both lead to interviews. It has been a great tool.

USER QUOTE

Thank you.

Raven is an active, evolving product—design work is ongoing. For the latest updates or to learn more, feel free to reach out to me below at:cristianvaldesdesign@gmail.com.