How qaaph-zyld/chatbots works
Project type and purpose
• This is a full-stack, modular “Customizable Chatbots Platform.”
• It provides a React-based web UI, an embeddable web widget, and a rich Node.js/Express backend with chat-engine and NLP modules, persistent storage, analytics, and multi-channel integrations.
File structure
• Top-level “src/” contains:
• api/ (Express controllers, routes, Swagger docs)
• client/ (React SPA under src/client)
• bot/ & nlp/ (chatbot core: engine implementations, NLP processors)
• services/ (business-logic layer)
• database/ (Mongoose connection + schemas)
• integrations/ (Slack, web, channel abstractions)
• auth/ & enterprise/security/ (authentication, RBAC, audit logging)
• analytics/ & monitoring/ (data-collection, Prometheus exporter)
• context/, workflow, templates, training (conversation flow & training pipelines)
• web-widget/ (standalone embeddable widget with its own build)
• Deployment artifacts: Dockerfile, docker-compose(.yml), kubernetes/ (Helm chart), GitHub workflows.
• Static/demo sites under public/, demos/, community/forum/, website/.
• Scripts/ orchestrates CI/CD, testing, security audits, backups.
README for additional insights
• Core features: multiple AI engines (Botpress, Hugging Face), conversation templates, NLP, multi-channel.
• Advanced: context management, knowledge-base, analytics dashboard, training UI, RBAC.
• Tech stack: Node.js ≥16, Express, React, MongoDB, Docker, Kubernetes, Prometheus, Jest, Playwright.
• Architecture principles: modular design, factory patterns, DI, environment-driven config.