fix: add limit param to prevent timeout

feat: return all results like original API

fix: match original approach, remove restrictive settings

fix: correct parquet prefetch param name

fix: reduce memory to 1GB, threads=1, no prefetch

fix: stable VIEW approach, no background loading

fix: background TABLE loading with indexes, persistent storage

fix: use VIEW with parquet filter pushdown for instant startup

fix: background loading with persistent storage to avoid 30min timeout

optimized: table instead of view, indexes, multi-field search

Update app.py

Update Dockerfile

Update requirements.txt

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update requirements.txt

Update Dockerfile

Update README.md

Create Dockerfile

Create requirements.txt

Create app.py

initial commit