Kairon is an open-source, no-code platform for building and deploying conversational AI digital assistants. It targets teams and individuals seeking an easy-to-use interface for managing the end-to-end lifecycle of AI assistants, from data pre-processing and training to deployment and monitoring, with a focus on proactive user support.

How It Works

Kairon builds upon the Rasa framework, extending its capabilities with tools for data augmentation, knowledge graph generation, and intent creation. It offers a low-code/no-code web interface for functional users and features proactive intervention through real-time struggle analytics. The platform supports omni-channel, multi-lingual deployment and includes a chat client designer and analytics module for performance insights.

Quick Start & Requirements

• Docker Deployment: Requires Docker and Docker Compose.
  • Clone the repository.
  • Modify docker/docker-compose.yml to set server to your public IP and optionally configure Google Analytics and SECRET_KEY.
  • Run cd kairon/docker && docker-compose up -d.
  • Access at http://localhost/ with demo credentials: test@demo.in / Changeit@123.
• Development Installation: Requires Python 3.10 and MongoDB 4.0+.
  • git clone https://github.com/digiteinfotech/kairon.git
  • Follow provided links for virtual environment setup.
  • Run setup.bat (Windows) or sh ./setup.sh (Linux).
  • Start augmentation services: python -m uvicorn augmentation.paraphrase.server:app --host 0.0.0.0
  • Start trainer-api services: python -m uvicorn kairon.api.app.main:app --host 0.0.0.0 --port 8080

Highlighted Details

• Leverages Visual LLM Chaining for proactive digital assistants.
• Features a "telemetry" system for monitoring user interactions and proactive intervention.
• Includes a "realtime struggle analytics" module to identify and address user friction.
• Offers a "chat client designer" for customizable user interfaces.

Maintenance & Community

The project is maintained and supported by NimbleWork Inc. Contributions are welcomed. Links to community channels are not explicitly provided in the README.

Licensing & Compatibility

Licensed under the Apache License, Version 2.0. This license is permissive and generally compatible with commercial and closed-source applications.

Limitations & Caveats

The README notes that documentation is still being fleshed out. A dependency issue with tensorflow-text on Windows is mentioned, requiring manual removal from requirements.txt. The project's current NLP focus is on Knowledge Graphs.