MockIT
AI-powered technical interview platform conducting real-time voice interviews with "Moki," an advanced AI interviewer that revolutionizes the hiring process.
See MockIT in Action
Watch a live technical interview conducted by Moki, our AI interviewer
Overview
MockIT transforms technical hiring through natural, conversational AI interviews. The platform conducts real-time voice interviews across multiple programming languages, providing detailed assessments and automated reports for companies.
As Head of AI & Co-founder, I architected the entire AI conversation engine, real-time audio processing system, and intelligent assessment framework.
See the live platform (in English and Polish).
Technical Implementation
AI Conversation Engine:
- Multi-layered prompt engineering with 15+ conversation actions
- Multiple AI providers (OpenAI GPT-4o, Claude 3.5, Groq, DeepSeek)
- Context-aware responses with dynamic action selection
- Built-in cheating detection and behavioral analysis
Real-Time Voice Processing:
- WebSocket bi-directional audio streaming (WAV input, MPEG output)
- ElevenLabs integration for natural voice synthesis
- Sub-second latency audio processing pipeline
- Advanced speech recognition with context corrections
Backend Architecture:
- Python 3.12 with FastAPI framework
- PostgreSQL with Alembic migrations (25+ entity models)
- Docker containerization with production deployment
- Async architecture handling concurrent interviews
Key Features
- Multi-dimensional assessment system (technical + communication skills)
- Support for 100+ technologies (Python, JavaScript, Java, DevOps, AI/ML, etc.)
- Automated PDF report generation with detailed analysis
- Traffit HR platform integration for workflow automation
- Multi-tenant system with role-based access control
- Real-time interview difficulty adjustment
- Multi-language support (English and Polish)
Technologies Used
Python 3.12FastAPIPostgreSQLDockerWebSocketsOpenAIAnthropicElevenLabsAWS S3SQLAlchemyPydanticPytestSentry