Processing Files

[New in version 3.0]

Private AI file processing automatically extracts text from files, documents and images and redacts it to return a redacted document in the same format as the original file.

How Does It Work?

Private AI support for file redaction comes in unified endpoints which works with URIs or base64-encoded files: /v3/process/files/uri and /v3/process/files/base64.

The uri endpoint gives you the ease of pointing to a file on a mounted drive and redact it without having to first read the file in memory or worry about saving the redacted content. The redacted contents are automatically saved at a user-specified location with the .redacted suffix added to the original name. For example, the uri endpoint will access the file /some/path/my-doc.pdf, it will redact it and create a file my-doc.redacted.pdf with the redacted contents at the location specified by the user. When using the uri endpoint, the file extension is used to determine the file type.

The base64 endpoint, on the other hand, is ideal if you want to save the redacted file yourself. To use the base64 endpoint, you first need to read the file in memory, encode its content with base64, and send it to the base64 endpoint. You also need to pass the MIME type of the file as a hint to the file processing pipeline.

The tables below summarize the supported file types, extensions and MIME types for both endpoints.

Documents

Images

Audio

Diving Deeper

Processing files with the uri endpoint

To process files with the /v3/process/files/uri endpoint you are required to mount a volume when starting the container.

docker run --rm -v <full path to your license.json file>:/app/license/license.json \
-v <full path to files>:<path in container> \
-p 8080:8080 -it crprivateaiprod.azurecr.io/deid:<version>

In addition, the service requires access to a folder where the redacted files will be stored. This is done with the PAI_OUTPUT_FILE_DIR environment variable. This variable must point to a folder that is mounted into the container or an existing subfolder to a mounted folder.

docker run --rm -v <full path to your license.json file>:/app/license/license.json \
-v <full path to files>:<path in container> \
-e PAI_OUTPUT_FILE_DIR=<path to mounted folder in container> \
-p 8080:8080 -it crprivateaiprod.azurecr.io/deid:<version>

This is an example of a command mounting a files folder in the admin home folder.

docker run --rm -v /home/admin/license.json:/app/license/license.json \
-v /home/admin/files:/files \
-e PAI_OUTPUT_FILE_DIR=/files/output \
-p 8080:8080 -it crprivateaiprod.azurecr.io/deid:3.1.0-cpu

Once the container is running with the above command, you can redact files with:

 curl --request POST --url 'http://localhost:8080/v3/process/files/uri' \
   -H 'Content-Type: application/json'  -d '{"uri": "/files/sample.pdf"}'

Upon successful completion, the above command will save the redacted file under /home/admin/files/output/sample.redacted.pdf.

docker run --rm -v /home/admin/license.json:/app/license/license.json \
-e PAI_OUTPUT_FILE_DIR=/files/output \
-v /home/admin/files:/files \
--user $(id -u):$(id -u) \
-p 8080:8080 -it crprivateaiprod.azurecr.io/deid:3.1.0-cpu

Checkout the API reference for more details on the uri endpoint.

Processing files with the base64 endpoint

When using the /v3/process/files/base64 endpoint, there is no need to mount a folder into the container.

docker run --rm -v <full path to your license.json file>:/app/license/license.json \
-p 8080:8080 -it crprivateaiprod.azurecr.io/deid:<version>

The file is first read into memory to encode its contents then the encoded contents are passed to the file processing endpoint. On linux, this can be done with the base64 shell command. Assuming you have the file sample.pdf saved in the current folder:

echo '{"file": {"data": "'$(base64 -w 0 sample.pdf)'", \
"content_type": "application/pdf"}}' \
| curl --request POST --url 'http://localhost:8080/v3/process/files/base64' \
-H 'Content-Type: application/json'  -d @-

This command will redact the file contents and return the redacted document as a base64-encoded string.

Checkout the API reference for more details on the base64 endpoint.

Limitations

Private AI is constantly improving the file processing support in every releases. These are the current limitations: