Local LLMs, RAG systems, and agent workflows running on dedicated infrastructure in Germany. Fully self-hosted, no data leaves the server.
A personal research project by Aymen Mastouri. Not affiliated with any employer or client.
Identity & Access Management. Central SSO for all lab services.
auth.aymenmastouri.ioChat interface with RAG support. Qdrant + nomic-embed-text for document Q&A.
chat.aymenmastouri.ioOpenAI-compatible API proxy. Routes to local Ollama models.
llm.aymenmastouri.ioVector database for semantic search. Connected to nomic-embed-text via LiteLLM.
qdrant.aymenmastouri.ioLocal LLM runtime. qwen2.5:3b + nomic-embed-text + llama3.2:3b.
internal · :11434Workflow automation for AI agent pipelines.
Document parsing and extraction for RAG ingestion.
Visual low-code builder for AI agents and chains.