WhyHire: from concept to launch.
How we take WhyHire from a strong idea to a live, two-sided hiring platform that master's graduates — and the companies competing to hire them — actually want to use.
This isn't only a plan — Phase 0 is already live.
The core experience is built and shipping. Swipe it, don't just read about it.
The opportunity
WhyHire is swipe-to-match hiring for master's-level graduates — a Tinder × LinkedIn × Indeed mashup where the candidate drives the match.
A graduate swipes through curated, master's-level roles. A right-swipe instantly sends that company's recruiter a Candidate Score Card — and the recruiter answers with one of four moves: invite to interview, request five video answers, send a personality profile, or pass. Both sides are scored, so fast and fair behavior wins visibility. No more applications vanishing into a black hole.
The thesis: win the wedge, not the world
The honest truth about marketplaces: the concept is strong, but what kills products like this is never the technology — it's the two-sided cold-start. No jobs means no candidates; no candidates means no jobs.
A swipe app with no jobs is dead on arrival. WhyHire is born with a built-in supply of both sides — a graduating class, and the employers who already recruit from it.
So we don't launch broadly. We win one wedge at a time:
- One beachhead school — start with the founder's own UC Irvine (Merage) network: warm intros, a captive graduating cohort, and a career-services office that already needs this.
- One or two grad fields and one metro — concentrate supply and demand so matches actually happen.
- Concierge the first matches by hand. Behind the curtain, we nudge both sides so the early experience feels alive — the same playbook Airbnb, DoorDash, and Tinder used to start.
The university angle isn't a feature. It's our distribution channel, our candidate supply, and a future revenue line — the one thing LinkedIn and Indeed can't easily copy at the graduating-cohort level.
What's already built
Phase 0 is live at whyhire.vercel.app — a polished landing page, a three-audience waitlist, and a fully playable prototype of the core loop.
It demonstrates the exact “magic moment” that makes WhyHire different — candidate-led matching and a recruiter who answers with intent, not silence. It's the proof you can put in front of a school, an employer, or a co-founder this week.
The product: two surfaces, one loop
| Surface | User | Form | Why |
|---|---|---|---|
| Candidate app | Master's grads | Mobile-first | Swiping is a phone gesture |
| Recruiter console | Employers | Web dashboard | Score Cards, pipeline, posting, analytics |
| Ops / admin | The team | Web (internal) | Concierge matching, moderation, metrics |
| Cohort view | Universities | Web (later) | Placement dashboards — a future paid seat |
MVP scope — ruthless
The first version serves one magic moment and nothing else. Everything that doesn't make the loop feel fast, fair, and a little addictive waits.
✓ In the MVP
- Auth + candidate profile (the Score Card)
- Recruiter job posting
- Swipe + match
- The four recruiter actions
- Async video via Mux/Cloudflare (don't build video)
- Templated/licensed personality step
- Rule-based matching
- Event logging on every swipe & decision
○ Deferred (V1+)
- Machine-learning matching
- Native iOS / Android apps
- Payments / billing
- Public recruiter grades, badges, leaderboards
- ATS integrations
- University dashboards
Matching: rules now, ML later
“Machine-learning matching” is a V3 capability, not an MVP one — treating it as MVP is how startups stall for a year.
We start deterministic: degree level, field, location, work authorization, salary band, must-have skills. The thing that earns real ML is the data we collect from day one — every swipe, every recruiter decision, every outcome. That labeled data is the only thing that makes a matching model real later.
Instrument first, model second. Each swipe and decision is a training label. By the time we're ready for ML, we'll own a proprietary dataset of what great grad-to-role matches actually look like — defensible, and impossible to buy.
The technology
A pragmatic, proven stack chosen for speed and low cost — the same tools already running the live Phase 0.
| Layer | Choice | Why |
|---|---|---|
| Web (recruiter, marketing) | Next.js on Vercel | Fast, serverless, already in use |
| Candidate mobile | PWA → React Native (Expo) | Ship fast, go native once validated |
| Backend / database | Node/TS + Postgres (Neon) | Scales cleanly, low ops |
| Auth | Clerk / Auth0 | Don't hand-roll auth |
| Video | Mux / Cloudflare Stream | Recording & playback as a service |
| Analytics / events | PostHog | The data flywheel |
| Payments (V1) | Stripe | When we turn revenue on |
The roadmap
Four phases, each with a gate we must clear before spending on the next. We're through the first one.
Validate
- Landing + waitlist
- Clickable demo
- Discovery calls
MVP loop
- Swipe → match
- Score Card + actions
- Rule-based match
V1 platform
- Recruiter console
- Video + personality
- Payments + analytics
Scale + ML
- ML matching
- iOS / Android apps
- University deals
The business model
We don't charge in the MVP — we earn the right first. Revenue turns on at V1, once employers feel the value.
- Featured & boosted postings — employer advertising on the highest-intent roles.
- Recruiting-solutions subscription — applicant tracking + employer-branding tools.
- University / career-services SaaS — the wedge becomes a recurring revenue line.
- Partner add-ons — résumé and assessment vendors, revenue-shared.
Legal & trust — built in, not bolted on
This is a hiring product, and that carries real obligations. Handled right, “bias-audited matching” becomes a selling point — not a liability.
- EEOC / adverse impact — any scoring or matching is audited for disparate impact.
- NYC Local Law 144 & Illinois AIVIA — bias audits and consent for automated decisions and AI video interviews.
- Data privacy — CCPA/GDPR, candidate PII, deletion rights.
- Recruiter scores stay private & aggregate early, to avoid fairness and defamation risk.
A short startup employment-law review before launch is non-negotiable — and cheap insurance.
The build — and what it's worth
The concept calls for a CTO, a CMO, and a CFO. Omniscient covers all of it — strategy, brand, marketing, and the full application — built end-to-end with an AI-accelerated engineering workflow (Claude, Codex, Cursor). No technical co-founder to recruit, no agency to manage. That's why the roadmap above is measured in weeks, not quarters.
What this would cost the traditional way
Current 2026 U.S. market rates to build WhyHire with an agency or a hired team:
A traditional MVP-to-V1 build runs ~$180K–$430K over 6–12 months; a fully scaled product, $400K–$900K+ — or a $500–800K-a-year in-house team. Shane is building WhyHire end-to-end as Esad's partner. The numbers above are the market benchmark, not an invoice — what an AI-accelerated workflow compresses into weeks, and what it's worth.
Benchmarks reflect 2026 U.S. agency/contract rates (senior full-stack ≈ $85/hr; two-sided marketplace MVP ≈ $40–120K). Ranges, not quotes.
What we do next
Phase 0 is live. The next moves are cheap, fast, and de-risk everything that follows.
- Lock the wedge. One school + one or two grad fields + one metro. Default: UC Irvine (Merage) MBA, Orange County.
- Turn on the leads. Wire the live waitlist to a real CRM so every signup reaches us instantly.
- Run 10–15 discovery calls. Five employers who hire master's grads, two career-services offices, eight grad students — validate willingness before we build.
- Greenlight the build. On your word, Omniscient starts Phase 1 — the MVP loop — and ships it in about a week.