An affidavit, PoA, Deeds of Trust, etc. - many documents are required from customers to start any legal proceeding. In this blog, we will show how you can automate the process of collecting such documents.

We will use the same trick as we did for supply chain documentation and extract the document header. Then we will assign a label to a document to identify it inside our board or a project. The final step would be to extract metadata (client's name, address, phone, etc.) from the document. 

The final result will look like this:

NOTE: In this demo, we will store the documents in Dropbox. However, if you use another cloud storage, the procedure will look the same.

For our processing queue to work we need a model that analyses documents and detects it's type. Let's create one - first, let's to go to A.I. models section in Bitskout and press Add:

Then let's choose the model type - Data Extraction and choose Extract Data from Documents.

Then we need to load the template and select the area - we will grab the data from top of the document because most of the documents today print their header (e.g., "Affidavit") at the top of the page. 

Once the area is selected, we can press Apply and have our model in the list.

Next, we will add a custom output that will setup labels. Go to Outputs -> Label Mapping and create a new output with the following configuration:

Then we create a workflow with all the data and we can use it in our monday.com board. 

To collect information from client you can use forms like this. 

Then you can add a Bitskout recipe and use the workflow that you've created.

There are multiple ways to setup the incoming flow and in the the next post we will share how to setup the whole end-to-end process with Integromat, JotForms, Dropbox and monday.com.