Hoovik: Open-Source AI Meeting Platform with Emotion Analysis
Hoovik delivers a complete solution for organizing intelligent video meetings, featuring real-time emotion analysis and transcription. Developed as open source, this platform leverages a distributed architecture to ensure high performance and scalability.
A modular and distributed architecture
Hoovik stands out with its modular approach, composed of four core services: a React frontend, a Node.js backend, a Python transcription service, and a Python-based emotion analysis module. Each component operates independently while communicating via standardized protocols.
The Node.js backend, built on Express and Socket.IO, handles authentication, room creation, and WebRTC signaling. It uses MongoDB for data storage and Redis for shared state management across multiple instances. This setup enables real-time synchronization of interactions, crucial for smooth meetings.
AI features designed for productivity
Beyond standard video conferencing, Hoovik integrates multi-speaker transcription enhanced with real-time emotion analysis. Meetings can be automatically summarized, while an augmented search system (RAG) extracts precise insights from transcriptions. A dedicated interface even streamlines access and approval of meeting minutes.
Two interactive demos are available online for hands-on exploration. The source code is also accessible on GitHub for developers looking to customize the platform to their needs.
🔗 Live Demo | Interactive Demo | GitHub
Source: DEV Community. Editorial synthesis assisted by AI — TechnoExpress.