Inspiration
In Lagos, 70% of tenants sign tenancy agreements they don't understand. Last year, my friend in Yaba lost ₦800,000 to a single clause: "non-refundable, guarantor must forfeit property if tenant defaults, 15% monthly interest on late payment." No lawyer, no explanation in Pidgin. She signed because she needed a house.
Nigeria has 40 million renters. Most can't afford lawyers. Laws like the Lagos Tenancy Law 2011, Stamp Duties Act, and CBN's 24% interest cap exist to protect them — but who explains it in plain language? We built LexGuardian NG — Naija Legal Guardian — to be that protector.
What it does
LexGuardian NG is an AI-powered contract guardian that analyzes any tenancy or loan agreement in <1 second using 4 specialized agents:
🔍 Agent 1 - Clause Detector: Finds critical clauses (rent, guarantor, forfeiture, interest, termination). ⚖️ Agent 2 - Compliance Checker: Cross-checks against Nigerian law — Stamp Duties Act, Lagos Tenancy Law 2011 (6 months notice required), CBN guideline (interest cannot exceed 24% annually). 📊 Agent 3 - Risk Scorer: Scores 0-100. Our test agreement scored 85/100 HIGH RISK - "GBEGE DON SUP, No sign am!" 🗣️ Agent 4 - Pidgin Explainer: Translates legal jargon to simple English + Pidgin: "Many kasala inside this paper. Guarantor fit lose him house. You need lawyer before you sign."
Users paste contract -> Click "Analyze with 4 Agents" -> Get risk score, flags, and bilingual explanation. Tagline: "No sign wetin you no understand!"
How we built it
We built it on my HP laptop inside DevSwarmProjects/Lexguardian-ng with Vite + React.
Architecture:
src/agents/index.js -> 4 functions: detectClauses(text), checkCompliance(text), analyzeContract(), explainPidgin()
src/components/RiskScore.jsx -> visual 0-100 gauge
src/App.jsx -> main UI with English/Pidgin toggle
Workflow: Text -> Clause detection (regex + keywords) -> Law compliance check (Nigerian law DB) -> Risk scoring -> Bilingual explanation.
We ran npm install && npm run dev -> Live at http://localhost:3000 -> Pushed 2435 objects to GitHub at Josephine2016/Lexguardian-ng. Full DevSwarm project, not desktop folder!
Challenges we ran into
- PowerShell quoting hell: Our first
Set-Content src\App.jsxgot stuck in PowerShell because@" "@here-strings broke — we learned to useCtrl+Cand simple writes. - Respectful Pidgin: Translating "unlimited guarantor liability" to Pidgin without sounding rude — "Guarantor fit lose him property" hits harder and is respectful.
- Making law simple: How do you explain Stamp Duties Act to a market woman in Yaba? We used short sentences and real consequences.
Accomplishments that we're proud of
- Live demo detecting 85/100 HIGH RISK on exploitative tenancy agreement in <1 second
- Bilingual English/Pidgin toggle — first legal tech in Nigeria to do this
- Successfully pushed to GitHub:
https://github.com/Josephine2016/Lexguardian-ngwith [new branch] main -> main - Works on HP laptop, offline-first logic, no API keys needed
- Built from "how do I locate folder" to deployed 4-agent AI app in one afternoon!
What we learned
We learned DevSwarm project structure, Nigerian legal tech gap is huge, and that "GBEGE DON SUP" is more powerful than "HIGH RISK" for real users. We learned to build for low-literacy users — not just lawyers.
What's next for LexGuardian NG
- Mobile App + WhatsApp Bot: Paste contract via WhatsApp to +234 number
- Lawyer Marketplace: For high-risk (80+) cases, connect to verified Lagos lawyers
- CAC Integration: Verify if landlord/company is registered
- Loan Agreements: Expand beyond tenancy to micro-loan sharks charging 15% monthly
Built With: React, Vite, JavaScript, DevSwarm, Nigerian Law, Pidgin NLP Team: Josephine Chukwu-Onwuchuruba, Lagos, Nigeria
What it does
How we built it
Challenges we ran into
Accomplishments that we're proud of
What we learned
What's next for LexGuardian-NG
Built With
- devswarm
- javascript
- natural-language-processing
- nigerianlaw
- react
- vite
Log in or sign up for Devpost to join the conversation.