Resource
Processing Thousands of Images in RAM
How we built a privacy-first bulk watermarking engine using in-memory processing.

Skip the Hard Drive
Most online image tools upload your files to a server, write them to a hard drive, process them, and leave them sitting there indefinitely. When you are dealing with unreleased client assets or proprietary product catalogs, that is a massive security risk.
Keep it Fast and Private
By utilizing Python’s io.BytesIO, our engine bypasses the disk entirely. We read, watermark, and zip your files in temporary server RAM. Once the download is complete, the memory is dumped, and your raw images vanish.
# Processing files directly in memory
zip_buffer = io.BytesIO()
with zipfile.ZipFile(zip_buffer, "w", zipfile.ZIP_DEFLATED) as zip_file:
img_byte_arr = io.BytesIO()
final_img.save(img_byte_arr, format='JPEG', quality=90)
zip_file.writestr(f"{filename}_wm.jpg", img_byte_arr.getvalue())