Fix: Clean formatted answers without LLM - proper summarization and readable output"

Fix search.py: add DB fallback when FAISS unavailable

Fix chat.py: pass db session to RAG for DB fallback search"

Fix RAG: fallback to DB text search when FAISS index unavailable"

Replace worker.py with a no-op stub (processing now in files.py)"

Remove celery from requirements (using BackgroundTasks instead)"

Simplify start.sh: no Celery needed, processing is inline via BackgroundTasks

Update frontend: real-time upload progress bar and processing status"

Rewrite files.py: inline processing with real-time SSE status updates

Remove jinja2 dep (using HTMLResponse directly now)

Fix template rendering - use new Starlette TemplateResponse API"

Fix start.sh: remove set -e, handle Redis failure gracefully

Fix Dockerfile: add user, fix permissions for HF Spaces

Fix files.py: handle Celery dispatch failure gracefully

Fix config: ensure data dir exists for SQLite

Fix app/worker.py: correct celery app name for CLI invocation

Fix app/main.py: handle missing Redis gracefully, add error handling

Fix startup: single worker, handle Redis failure gracefully, fix Celery import

Add templates/index.html - Full SPA frontend

Add static/style.css

Add app/main.py and templates/index.html

Add app/worker.py - Celery background processing

Add app/rag.py - RAG pipeline

Add app/search.py

Add app/chat.py, app/search.py, app/rag.py, app/worker.py, app/main.py

Add app/files.py

Add app/auth.py, app/files.py, app/chat.py, app/search.py

Add app/db.py - database models

Add app/config.py

Add app/__init__.py

Add start.sh

Add requirements.txt

Add Dockerfile for HF Spaces

Clean slate - rebuild with simpler architecture for Spaces

Add frontend package.json

Add backend requirements and core files

Add Dockerfile for Space deployment

Update ML Intern artifact metadata

Initial Space setup with README

Update ML Intern artifact metadata

initial commit