Zepyoc – Audio-Based Hiring Platform
Project Details
Project Overview
Zepyoc was built to address a core inefficiency in modern hiring: resumes rarely reflect how a candidate actually communicates or thinks. By shifting the first screening step to audio-based responses, Zepyoc allows recruiters to evaluate candidates more holistically while significantly reducing screening time.
Mission Statement
The mission was to create a production-ready hiring platform that feels fast, fair, and human. The system needed to support structured audio workflows, scale across multiple job postings, and provide recruiters with a clear, efficient review experience.
Philosophy
Zepyoc is built around the idea that hiring should be human-first. Voice captures nuance, intent, and communication ability far better than text. The platform uses technology to remove friction, not replace judgment.
Design Principles
- Audio-first, not resume-first
- Structured workflows over free-form noise
- Fast screening without sacrificing context
- Simple interfaces for busy recruiters
- Scalable systems ready for AI augmentation
Technical Architecture & Innovations
Core Technology Stack
- - Next.js 14 – SaaS frontend and server components
- - TypeScript – Type-safe product development
- - Node.js – Core application backend
- - Python & FastAPI – Audio processing and AI services
- - Supabase – Authentication, database, and file storage
- - PostgreSQL – Relational data and application state
- - Cloud storage – Secure audio file management
Development Phases
- • Authentication and user roles
- • Job posting and application schema
- • Audio recording and upload system
- • In-browser audio recording
- • Audio file storage and streaming
- • Candidate response playback
- • Error handling and retries
- • Performance optimizations
- • Responsive UI refinements
- • Production deployment
- • Monitoring and logging
- • Initial onboarding flows