Consumer · Career AI agent · Live product · free plan · public demo
Your next opportunity is out there. Wonder finds it.
WonderJobs scans live sources, removes duplicates, scores every posting against a candidate's Career DNA and drafts tailored materials — then hands the application to the human. It never applies on anyone's behalf.
- Platform
- Web · PWA · desktop, tablet, mobile
- AI
- Anthropic Claude (platform) · Anthropic / OpenAI / Gemini BYOK
- Data
- Supabase PostgreSQL, dedicated schema, RLS, service-role only
- Pricing
- Free today · Pro in development
The problem
Job searching is a full-time job. It shouldn't be.
Candidates spend evenings re-typing the same details into the same forms, scanning the same boards and guessing whether a role is worth it.
- 1
Repetition without signal
The same listing appears on six boards. The same résumé is rewritten for the tenth time. None of it tells the candidate where they actually fit.
- 2
Opaque matching
Boards rank by recency or by who paid. A candidate cannot see why a role appeared, or how strong the fit really is.
- 3
Automation that overreaches
Auto-apply tools submit on the candidate's behalf, damaging reputation with employers and violating platform terms.
- 4
Ghost listings and noise
Stale, duplicate and low-quality postings waste the scarcest resource a job seeker has: attention.
The product
Search, analyse, prepare, track — with the candidate in control.
A run is an explicit twelve-stage state machine: understand the profile, search sources, remove duplicates, score fit, check quality, prioritise, prepare materials, wait for review, hand off, track and learn.
01
Career DNA
Goal, level, years, rated skills, industries, locations, minimum salary, strengths and growth areas. Every real posting is scored against it — and rescored when it changes.
02
Explainable fit
Strong Opportunity (82+), Worth Considering (68+), Stretch (55+). Every score explains itself per dimension; quality signals use confidence language, never 'ghost job' claims.
03
Application prep, human hand-off
Tailored résumé, cover letter and screening answers, with version compare and restore. The candidate reviews, edits and submits. Every external action is audited.
04
Your AI, your keys
WonderJobs AI (Anthropic Claude) by default, or bring an Anthropic, OpenAI or Gemini key. Keys are AES-256-GCM encrypted, used only server-side, never returned.
How it works
From signal to outcome.
Step 1
Search
Company career sites (Greenhouse, Lever, Ashby), Remotive, Jobicy, Remote OK, Himalayas, Arbeitnow and Adzuna India — live, server-side, cached 15 minutes.
Step 2
Analyse
A deterministic normaliser derives skills, seniority, industry, work mode and salary from each posting's own text, then scores it.
Step 3
Prepare
Materials are drafted per role and queued for review. Automation levels — Assist, Guided, Autonomous, Continuous — gate every capability.
Step 4
Track
Applications, follow-ups and interviews build a timeline, a calendar feed and insights that improve the next run.
Product tour
Desktop and mobile, one product.
Captured from the deployed product. Some views show demonstration data.
Scroll-linked hero on the product site
Jobs view — every match explains itself
A run's live stages on mobile
Applications tracker
Career DNA — the profile every score is measured against
Installable PWA
Search · Analyse · Prepare · Track
Insights derived from real runs and applications
Your AI, your keys
Market & why now
Every active job seeker, every few years.
Job search is universal, recurring and painful. Incumbent boards monetise employers; the candidate side is under-served by tools that either spam or overreach. Remote-first hiring has widened the field a single person must scan.
Who buys
- Knowledge workers searching remote and hybrid roles
- Career switchers who need fit explained, not assumed
- Students and early-career talent entering global markets
- Senior professionals managing a discreet, high-stakes search
Why now · 1
Structured public job feeds and company ATS boards make honest, real-data search possible without scraping walled gardens.
Why now · 2
Language models draft credible, tailored materials — but only trustworthy when paired with deterministic matching and human approval.
Why now · 3
BYOK lets candidates pay their own AI bill, keeping the free tier genuinely free and gross margin high.
Business model
Free to search. Pro for depth and automation.
The free plan is real and the demo runs without an account. Pro is the natural upgrade for scheduled runs, deeper coaching and higher limits; billing is designed but not yet connected, so today 'Upgrade' records interest. BYOK means heavy users can bring their own inference cost.
Free
Search and prepare
- Real sources, explainable matching
- Application materials and tracker
- WonderJobs AI or bring your own key
- Installable on any device
Pro
In development
- Scheduled runs with a server-side scheduler
- Deeper résumé coaching and mock interviews
- Higher usage limits
- Priority sources as they land
Defensibility
What compounds.
Honesty as a feature
Only sources actually searched are shown. Illustrative scenarios are labelled. Wonder never submits on an employer's site. Trust compounds where competitors burn it.
Deterministic core, generative edge
Matching, ranking and quality checks are rule-based and unit-tested; language models only draft prose. Cheap, auditable and consistent.
Local-first, sync-second
State hydrates locally and syncs per tenant in batches — fast on a phone, resilient offline, cheap to serve.
Execution to date
Verifiable, not aspirational.
Pulled from the product's own public engineering trackers. Source code, trackers and live demos are available to serious parties on request.
Epics complete
11 / 15
Remaining four are partial, with operator steps or provider config outstanding
Live job sources
7 families
Greenhouse/Lever/Ashby boards across 20+ companies plus six aggregators
Accessibility
0 violations
axe-core audit across 18 pages, reduced motion emulated
Production
Verified after each push
Auth, demo, sources and per-user state checked on every release
Roadmap
What comes next.
Now
- Server-side scheduler so runs fire with the app closed
- Email delivery for follow-ups and thank-you notes
- Résumé import to pre-fill Career DNA
Next
- Pro plan billing
- Two-way calendar sync
- Push notifications
Later
- Additional regional sources
- Mock-interview AI and deeper coaching
- Platform admin for usage and key rotation
The ask