Case study — 0 to 1 product design
IamReady AI: From Campus Wellness App to AI Hiring Platform
A campus wellness app with flat navigation and no revenue path was rebuilt into an AI interview-readiness platform. countless decisions that carried the pivot, here are the most impactful ones
Role
Solo Product Designer, end to end -research, information architecture, UX/UI. Development team built and shipped what I designed.
Type
0 to 1 product. B2B SaaS with a B2C candidate side.
Platform
Several months of iterative development
Live at
60%
Reduction in time-to-hire reported by companies
3×
More qualified candidates surfaced vs. keyword-only ATS tools
700M+
External profiles searchable through Agent TAM
Candidate interview practice became the Conversational AI Interviewer, running structured, role-specific interviews on demand.
Recruiter feedback and scoring became semantic AI scoring, matching candidates by meaning (“revenue operations” to “sales ops”) instead of exact keywords.
The sprawling multi-role IA collapsed into an embedded careers page ATS and a searchable Talent Hub.
Sourcing and verification became Agent TAM, which searches 700M+ external profiles, and Agent LIO, which mines a company’s own past applicants before it sources externally.
Credit and package pricing experiments matured into three plans: Free, Growth ($150/month), and Enterprise ($600/month).
Onboarding, an afterthought in the original notes, became its own structured module, triggered the moment an offer is accepted.
Once upon a time
Two ideas bundled into one platform
The project started as two ideas bundled into one platform, One half was a mock interview coach: students would practice for real job interviews with AI-driven questions and feedback. The other half was a campus mental health and harassment reporting tool: students could talk to an AI chatbot, book a counsellor, and file complaints about bullying, discrimination, or assault, with a path for universities and even government bodies to monitor outcomes.
The landing page summed up the pitch: “Elevate Your Career and Well-Being.” Career prep and mental wellness sat side by side as equally weighted pillars, aimed at students under academic and emotional pressure.
Every day
Every design cycle added more surface area
There was a student flow, a counsellor flow, a university admin flow, a recruiter flow, and eventually a government oversight flow, all being sketched out in parallel. Priorities pulled in different directions: keep mental health and interview prep “equal,” but also build job applications, subscriptions, pricing tiers, and admin panels for multiple user types at once. Nothing was getting finished because everything was in motion.
One day
The mental health side hit a wall with the exact partners it needed most
Universities interested in the mock interview product explicitly did not want the complaint and harassment reporting piece bundled in. They didn’t want to share student data or lose control over sensitive cases. That was the forcing function: the platform could not be everything to everyone. It had to pick a problem people would actually pay to solve, and a user who could sign up without needing an institution’s permission first.
Because of that
The pivot wasn’t one decision, it was a series of smaller design problems, each forcing a rethink of a specific flow. Five that shaped the product most:
01
Splitting products at the IA level, not just the pitch level
When universities said they wanted interview practice but not the harassment-reporting and mental-health-data product bundled with it, because they didn’t want to own or be liable for that data, I separated the two at the account and navigation level rather than just repositioning the marketing. In the merged dashboard, mock interviews, job applications, and Mr Safe get equal visual weight, but a returning user never has to sign up again for Mr Safe specifically, so the wellness side stays low-friction while running on its own consent boundary. This is the decision that made the eventual hiring-platform focus possible.
02
Turning “questions and a score” into an actual interview simulation
Competitive research on micro1.ai’s AI interviewer forced a harder question: was this simulating a real interview, or just scoring answers to a fixed list? I pushed the design toward letting recruiters supply their own question bank for the AI to draw from, made difficulty selectable (easy, medium, hard), and branched the flow by interview type — general, a specific pasted job description, MBA admissions, or civil services exams — instead of running one generic script for everyone.
03
Progressive monetization instead of a hard paywall
Recruiters could post a job free and get up to 200 applicants, but AI screening and AI interviews cost extra. Rather than gating the whole flow behind a purchase, I designed the candidate list to check purchase state on click: already bought AI interview credits, see the video and score directly; haven’t bought, the same click opens the buy page instead of a dead end. Remaining credits show in a “how many left” popup rather than surfacing only once they run out.
04
Trust and consent before high-stakes actions
Once video interviews entered the picture, candidates needed to know upfront that the session was recorded and, once started, couldn’t be paused or restarted. I added a blocking confirmation screen before any assessment or company interview begins, stating the no-going-back policy and what happens to the video, so candidates consent knowingly instead of discovering the rules mid-interview.
05
Making semantic matching legible, not just accurate
Traditional ATS tools reject good candidates over exact-keyword misses — someone who wrote “revenue operations” instead of “sales ops,” for example. Rather than let the AI score candidates as a black-box number, I designed the match itself to show its work: each profile displays the specific skill pairs it matched (Revenue Ops to Sales Operations, Team Leadership to People Management) and how strong each pairing is, so a recruiter can see why the AI made the call. This shipped as the platform’s core differentiator and its most repeated praise point: one hiring manager said it “changed how we think about candidates. We stopped rejecting people for using different words.”
Until finally
A hiring platform, not a wellness platform
The scope that survived all of that iteration was narrower and sharper. Candidates sign up free (B2C): build a profile, upload a CV, practice interviews, and apply to jobs. Companies pay for AI-run recruiting operations (B2B SaaS): source, screen, interview, and onboard.
Nearly every earlier exploration found a home in what shipped as IamReady AI:
Ownership
I owned the research, information architecture, and every design decision on this project end to end, working alongside a development team that built what I designed. Customers have called out faster, more confident hiring decisions and semantic scoring that changed how they evaluate candidates who describe their experience differently than the job spec does.