Dreaming
01 / Saas Chat CMS
LLMCMS is a multi-user AI platform designed to simplify local LLM deployment with Retrieval-Augmented Generation (RAG), customizable AI workflows, and configurable response filtering. Originally built as an experimental project using LangChain, it evolved into a flexible platform that supports multiple users, isolated workspaces, and modular AI capabilities.
AI-generated responses are processed through a configurable filtering pipeline before reaching users. Administrators can define custom filtering rules to control and sanitize AI output based on project requirements.
Upload documents to build a searchable knowledge base. The system automatically processes files, generates embeddings, and retrieves relevant information to provide more accurate AI responses.
Expose AI capabilities to external applications through configurable APIs. Permissions can be managed to control which services are available for different integrations.
Monitor and understand AI usage through a centralized analytics dashboard. Track LLM usage, interaction modes, task status, and processing time to gain a clear overview of system activity and performance.
Easily configure AI models, prompts, and providers through a web-based interface without modifying the source code. Switch between local models and cloud-based LLM providers to suit different deployment needs.
Interact with AI using voice through speech recognition and real-time voice conversations, providing a more natural and hands-free user experience.
02 / Related Informations