Team Member Prompts
User types a question, uploads an image, or requests creative generation in the web chat.
Self-hosted LibreChat AI workspace orchestrated via LiteLLM proxy, connecting Google Vertex AI (Gemini 1.5 to 3.8 Flash, Nano Banana Pro), OpenRouter, and Replicate with mongo-data and meili-data persistence

30+ frontier LLMs, vision engines, and private vector search under one zero-leak cockpit.
Multi-provider routing across Vertex AI, OpenRouter, and Replicate with persistent mongo-data and meili-data search.
Frontier LLM and generative vision models unified under a single self-hosted ecosystem with zero vendor lock-in.
Single-provider AI tools fragmented user history, lacked multimodal image/video generation, and suffered sluggish search.
Engineered LiteLLM proxy bridge linking LibreChat to Vertex AI, OpenRouter, and Replicate with dual Mongo & Meili storage.
CASE STUDY ROLE: AI INFRASTRUCTURE ARCHITECT & LITELLM GATEWAY INTEGRATOR
Xavier took the open-source LibreChat platform and turned it into an enterprise-grade AI hub.
He introduced LiteLLM as the central smart router to seamlessly connect Google Vertex AI, OpenRouter, and Replicate, then configured a dual-database layer with MongoDB for chat persistence and Meilisearch for instant search.
User types a question, uploads an image, or requests creative generation in the web chat.
Selects the optimal model (Gemini for speed, Claude for writing, DeepSeek for code, Replicate for visuals).
Securely queries Vertex AI, OpenRouter, or Replicate without exposing enterprise API keys.
Answers stream to the screen while mongo-data saves the chat and meili-data indexes it for instant search.