TrustHire AI — Verify Before You Trust
The Inspiration
Every day across Africa and the world, thousands of job seekers lose money, personal data, and hope to fake recruiters and fraudulent job offers. In Kenya alone, WhatsApp and Telegram are flooded with "remote job" scams promising $5,000/month for simple data entry — with an "onboarding fee" required upfront.
I built TrustHire AI because I've seen this happen to real people. There was no simple, accessible tool to verify an opportunity before it was too late. That gap needed to be filled.
What It Does
TrustHire AI is a full-stack web application that helps job seekers, freelancers, students, and remote workers instantly verify whether a job offer, recruiter, or business opportunity is legitimate or a scam.
Core features:
- AI Risk Scoring Engine — paste job details, get a 0–100% fraud score instantly
- Signal Detection — flags personal emails, upfront payments, unrealistic salaries
- Community Scam Reports — searchable database of reported scams
- Recruiter Verification — check if a recruiter's identity holds up
- Premium Dashboard — track all your analyses in one place
How I Built It
I built TrustHire AI entirely using MeDo — no manual coding required.
I structured my MeDo conversations in 4 deliberate phases:
Phase 1 — Architecture I described the full app structure: all 5 pages, navigation, sidebar, form fields, and core functionality in a single detailed prompt.
Phase 2 — Visual Design I prompted MeDo to apply a premium glassmorphism design system — frosted glass cards, deep navy-to-purple gradients, soft glow effects, and smooth animations to match a production-grade SaaS product.
Phase 3 — Risk Logic I defined the weighted scoring rules in plain language and MeDo translated them into working logic:
| Signal | Risk Points Added |
|---|---|
| Personal email (Gmail/Yahoo) | +25 |
| Upfront payment requested | +35 |
| Unrealistic salary | +20 |
| No company website | +20 |
Phase 4 — Sample Data & Polish I added realistic Kenyan market context, sample scam reports, and micro-interactions — animated risk score circles, hover glows, and loading states.
Challenges I Faced
The biggest challenge was scope control. It's tempting to build everything — browser extensions, heatmaps, email analyzers. I had to discipline myself to build 5 pages exceptionally well rather than 10 pages poorly.
The second challenge was making mock data feel real. I spent time crafting sample scam reports with authentic details — real Kenyan companies, realistic salary figures, and genuine scam patterns — so the platform feels fully operational during a demo.
💡 What I Learned
- MeDo is most powerful when you give it structured, phased prompts rather than one giant instruction
- A focused MVP with one exceptional feature beats a bloated app every time
- No-code doesn't mean low quality — TrustHire AI looks and feels like a funded startup product
What's Next
- Integrate real AI API for live job description analysis
- Connect to scam databases and WHOIS domain lookup APIs
- Launch browser extension for real-time job board scanning
- Partner with African job platforms to protect their users
Built with MeDo. Protecting opportunity seekers, one scan at a time.
Built With
- ai-risk-engine
- css
- glassmorphism
- javascript
- medo
- no-code

Log in or sign up for Devpost to join the conversation.