billsense-pdf
SECRET_SET=yes DPI=200 MAX_PAGES=20
pymupdf=yes pdfplumber=yes tesseract=yes easyocr=disabled opencv=yes preprocess=disabled ocr_concurrency=4
POST /extract-text — digital PDF text + tables (pdfplumber)
POST /rasterize — multipart file + output_dir (optional)
POST /extract — multipart file (raster + tesseract)
POST /ocr-easy — page images (multipart files[]); spawns subprocess, then exits