Which AI services does the ai-core skill support?

It provides configurations and best practices for LiteLLM (proxy), vLLM and Ollama (inference), LangFuse (observability), Qdrant (vector DB), and n8n (automation).

Does it include monitoring for AI responses?

Yes, it integrates LangFuse observability into the stack, allowing you to trace generations, monitor latency, and score model outputs for accuracy.

Can I use this skill for GPU-accelerated inference?

Yes, it includes specific Docker configurations for vLLM with AMD/ROCm support and optimized GPU memory utilization settings.

How does it handle model routing?

It utilizes LiteLLM to create a unified API endpoint that can route requests to local models (Ollama/vLLM) or external providers like OpenAI and Anthropic with automatic fallbacks.

AI Core Command Center

Name: AI Core Command Center
Author: Lobbi-Docs

byLobbi-Docs

Ciencia de Datos y ML

Deploys and manages a self-hosted AI infrastructure stack including LLM proxies, inference servers, vector databases, and observability tools.

Acerca de

The AI Core skill provides a comprehensive framework for orchestrating the Ahling Command Center's AI layer. It enables developers to implement a robust, local AI pipeline using LiteLLM for model routing, vLLM and Ollama for high-performance inference, Qdrant for vector search, and LangFuse for end-to-end observability. By providing standardized Docker Compose configurations and implementation patterns, this skill simplifies the complexity of managing multiple AI services, ensuring a scalable and observable environment for building LLM-powered applications and automated workflows.

Características Principales

Unified model routing and load balancing via LiteLLM proxy
High-throughput inference configuration for vLLM and Ollama
Comprehensive LLM observability and tracing with LangFuse
AI-driven workflow automation and event handling with n8n
Vector database management and embedding storage using Qdrant
0 GitHub stars

Casos de Uso

Setting up a private, self-hosted AI infrastructure for secure data processing
Building a multi-model gateway with automatic fallbacks and performance monitoring
Creating automated AI agents that interact with smart home or enterprise APIs via n8n

Acerca de

Características Principales

Unified model routing and load balancing via LiteLLM proxy
High-throughput inference configuration for vLLM and Ollama
Comprehensive LLM observability and tracing with LangFuse
AI-driven workflow automation and event handling with n8n
Vector database management and embedding storage using Qdrant
0 GitHub stars

Casos de Uso

Setting up a private, self-hosted AI infrastructure for secure data processing
Building a multi-model gateway with automatic fallbacks and performance monitoring
Creating automated AI agents that interact with smart home or enterprise APIs via n8n