fix(parser): implement explicit pdfplumber table extraction with markdown separators

chore: finalize Axiom RAG pgvector migration, DOCX parsing, and LLM reliability

fix:Changed chunks size

feat: implement table-aware chunking and increase RAG context window

feat: implement table-aware chunking and increase RAG context window

chore: completed global codebase rebranding to Axiom

fix: corrected get_supabase import path in leads.py

feat: enterprise production updates and Axiom rebranding

Updated Chat Interface and Admin Panel

Add user approval flow with admin notifications

Add rate limiter and tenant branding modules

Code quality, security, and cleanup improvements

Fix Scanned document text extraction

Fix admin authentication to use live Supabase query instead of JWT

feat: add all untracked files — frontend, core, routes

chore: remove Wikipedia LFS files — Business Mode only

fix: hf deployment crash

fix: add /admin/check endpoint and fix admin status check for business mode toggle

fix: disable google_auth router to fix HF deployment

Added auth, documents routes, updated frontend and core logic

Add profile page with name/password update and stats

Add streaming responses and multi-turn conversation

Fix: load pipeline in background thread at startup

Fix: skip BERTScore and local models in production

Fix: lazy load pipeline for Render deployment

Initial commit — Veritas RAG system