UPrise

A lightweight, reputation-first internship platform built for the African tech ecosystem — verified merit, not prestige.

🌍 The Mission

In Africa, the "Trust Gap" is the biggest hurdle to employment. Thousands of brilliant students from rural or less-known universities are ignored because HR managers rely on "prestige" rather than "proof."

UPrise levels the playing field. Using an Upwork-inspired Reputation Engine and AI-powered vetting, we replace biased CV-screening with verifiable merit.

🛠️ The Tech Stack

Built for speed, type-safety, and scalability:

• Framework: T3 Stack (Next.js 15, TypeScript, tRPC)
• Auth: Better Auth (Secure Session Management)
• Database: Turso (Edge-compatible SQLite)
• ORM: Prisma (Hardened Schema with Audit Logs)
• AI Engine: Google Gemini (Automated Grading & Plagiarism Detection)
• Styling: Tailwind CSS

🚀 Key Features

• AI-Vetted Skill Sprints: Candidates don't submit CVs; they solve real-world tasks graded instantly by AI.
• Upwork-Lite Reputation Engine: Features a Job Success Score (JSS) calculated from both public ratings and hidden private feedback.
• Blind Recruitment: Employers see scores and skills before names/photos to eliminate unconscious bias.
• AI Fraud Protection: Automated detection for unedited LLM-generated submissions.
• African Context: Built to be lightweight, offline-ready, and mobile-first for areas with fluctuating connectivity.

📊 Database Architecture

The system utilizes a denormalized Reputation Engine schema:

• Users & Profiles: Separated for performance and data integrity.
• Contracts: Tracks the full lifecycle from Hire to Completion.
• Dual-Review System: Captures public feedback and private "Trust Scores."
• AI Logs: Full auditability of AI tokens, latency, and raw responses.

Live demo

https://uprise-africa.vercel.app/

Folder structure (high-level)

• 📁 app — Next.js app routes and pages (app router)
• 📁 src
  • 🧩 _components — UI pages and dashboards
  • 🧭 admin, candidate, employer — role-specific flows
  • 🧰 components — shared UI components
  • 📚 lib — services and helpers (e.g. ai-service.ts, auth-client.ts)
  • 🧪 trpc — client/server tRPC setup
  • 🗄️ server — server adapters, API and auth helpers
• 📁 prisma — schema, migrations, and seed scripts
• 📁 public — static assets
• 📄 Project configs: package.json, tsconfig.json, next.config.js

How to run (local dev)

1. Clone: git clone <repo>
2. Install: npm install
3. Copy env: create a .env with DATABASE_URL and any AI keys (e.g. GEMINI_API_KEY).
4. Apply schema and seed (if needed):
   • npx prisma db push
   • If seeds exist: npx prisma db seed or node prisma/seed.ts
5. Start dev server: npm run dev

Using the app (short guide)

• Sign up / Sign in: use the onboarding flows under /onboarding for employers and candidates.
• Candidates: complete AI-vetted tasks to build skills and Job Success Score (JSS).
• Employers: create jobs, review blinded candidate scores, and invite for interviews.
• Admins: manage users, view AI logs, and moderate content from the admin dashboard.

Demo accounts

These demo accounts are created by prisma/seed.ts when you seed the database.

• Employer: employer@uprise.demo / 123456789
• Candidate: candidate@uprise.demo / 123456789

Run the seed script to recreate these accounts:

npx prisma db push
node prisma/seed.ts

Notes

• This README is intentionally brief. See the source files under src/ and prisma/ for implementation details.
• If you want, I can add a more detailed developer setup, environment variable list, or CI instructions.