Optimize RAG evaluation to unified single LLM call, increase timeouts, and handle 429 quota limits

fix: remove document_id constraint from config lookup in stream and chat endpoints

fix: resolve POST /api/chat/prepare 404 errors

chore: sync project updates

Major Chunking,Retrieval and Ranking Updates

Refactored the codebase and implemented stateless

Add auth reset flow, rate limiting, and frontend fixes

feat: add auth-gated 3-mode workflow with admin chroma controls

Implemented Compare feature and fixed other chunking strats

Fixed Chat function bugs and internal bugs related to UI

Implemented HF and Nvidia Embedding models and removed Gemini embedding

Implemented Fully Functioning Semantic and Fixed_size chunking strats with a change in RAG Pipeline

Major Bug Fixes and Implemented a Base RAG Workflow

Bug Fixes in the Main RAG PIPELINE

feat: add FastAPI routes — documents, chat, config, analysis