README.md

@@ -5,6 +5,19 @@ It contains file metadata compute, instant upload, calculating hashing, multi de
 
 ## Features
 
+Paperclip store and processing uploaded files with pipeline flow.
+When a user try to upload files. The file will store in local first for media processing.
+
+Then the server will publish a message into the message queue.
+And the background consumer will start dealing with the uploaded files.
+
+The background consumer will hash the file and merge the files with same hashcode.
+The background consumer will decode the image and generate ratio and read more info from image file too.
+
+After the processing done. The consumer will upload the file to the persistent storage like a s3 bucket and remove local cache.
+While the processing, the file record in database will marked to the temporary and load file from the temporary storage.
+When the processing done, the file record will be updated.
+
 ### Supported Destinations
 
 - Local filesystem