AI-powered mock interviews & data-driven hiring – built with Next 14, Supabase & GPT-4o
InterviewGrade is an advanced AI-driven mock interview platform designed to simulate real-world interviews and provide candidates with personalized feedback. It also offers tools for employers to streamline their hiring processes using data-driven insights.
- Candidates run real-time practice or mock interviews, get instant AI feedback and long-term skill analytics.
- Employers filter talent by verified soft-skill scores, unlock detailed reports with pay-as-you-go tokens.
- All grading is handled by GPT-4o + DistilRoBERTa sentiment, streamed back to the UI in seconds.
- AI-Powered Mock Interviews: Simulate real-world interviews with AI-generated questions and receive real-time feedback.
- Tailored Feedback: Get personalized feedback on key areas such as communication, problem-solving, and technical proficiency.
- Analytics Dashboard: Track progress over time and identify areas of improvement.
- Employer Interests: Candidates receive notifications when an employer unlocks there profile.
- Access to talented Candidate pool: Fine-grained search / filter on scores & resume
- Analytics Dashboard: Automated matchmaking candidate performance/resume based off candidate preferences to make better hiring decisions.
- Deeper Camdidate analytics: Employers can view deeper analytics about a desired candidate and pay one token(1 euro) to reach out to them.
- Stripe Integration: Recruiters can purchase different token bundles which will be used to unlock candidates of their choice.
| Layer | Main tech | Why |
|---|---|---|
| Frontend | Next 14 (App Router) · React 18 · Tailwind CSS · shadcn/ui | Edge streaming, typed RSC & Client Components |
| Backend | Next API Routes & Server Actions · Node 20 | Runs inside Vercel serverless / edge |
| Database | Supabase (Postgres + RLS) | SQL + JSONB, auth & storage bundled |
| AI | OpenAI GPT-4o, Whisper • Hugging Face DistilRoBERTa | Grading, feedback & sentiment |
| Payments | Stripe Checkout + webhooks | PCI-compliant token bundles |
| DevOps | Vercel, Docker (local Supabase) | Push-to-deploy, zero infra mgmt |
| Tooling | TypeScript • Zod • ESLint / Prettier • GitHub Actions | End-to-end type safety & CI |
The project is organized as follows:
Prereqs: Node ≥ 20, pnpm (or npm), Docker Desktop.
-
Clone the repository:
git clone https://github.com/jaynwanze/interviewgrade.git cd interviewgrade -
Install Dependecies
pnpm install
-
Set up your environment variables: Create a .env.local file and add your Supabase credentials and other environment variables:
NEXT_PUBLIC_SUPABASE_URL=your-supabase-url NEXT_PUBLIC_SUPABASE_ANON_KEY=your-supabase-anon-key SUPABASE_SERVICE_ROLE_KEY=… OPENAI_API_KEY=… STRIPE_SECRET_KEY=… STRIPE_WEBHOOK_SECRET=…
-
Download and start docker engine: Start Docker Desktop - Ensuring docker engine is running
-
Run the local db in docker env:
pnpm supabase start
-
Run the development server:
pnpm run dev
-
Open the application in your browser:
http://localhost:3000