GoonConverting PDF Documents to Text using AI
In today's video, we're talking about how you can convert PDF documents into text and then use that text for various purposes. For instance, let's say you have a contract and you want to pull out the line items. Traditionally, you'd probably have to do this manually, entering each item into a Google sheet, which can be a tedious and time-consuming task, especially if you have a ton of contracts to go through.
Introduction to Make.com
We're using a platform called Make.com to automate this process. If you're brand new to Make.com, there's a crash course available that takes you through everything you need to learn. The blueprints for this automation are available for free in the description, and all you have to do is download the JSON file, import it into your Make.com account, and you'll have this exact scenario ready to go within the next 30 seconds.
Setting Up the Workflow
To start the workflow, we need to set up a trigger. In this example, we're using the Google Drive trigger, so anytime a PDF enters the specified folder, the workflow will start. The first step is using OCR (Optical Character Recognition) to convert the PDF document into plain text.
Converting PDF to text using OCR is the first step in our workflow
Extracting Data from the Document
Once we have the text, the next step is to extract all the important information from the document. We're using an Open AI Chat GBT module to do this. Practically speaking, what's going on here is we're taking the text and asking the AI to pull out specific information like the invoice number, date, subtotal, and line items.
Using Chat GBT Module
In the Chat GBT module, we've created three particular messages. The first is a system message that gives context to the AI, telling it what it's meant to do. The second is an assistant message that structures the data in JSON format. The third is a user message where we're passing the actual text from the agreement into the AI.
Using the Chat GBT module to extract specific data from the document
Posting Data to Google Sheets
Once we have the extracted data, we use something called an iterator to go through each line item one by one and post it into a Google sheet. This way, we can automate the process of entering data from contracts into a spreadsheet.
The extracted data is then posted to Google Sheets using an iterator
Conclusion
In conclusion, this video has shown how you can automatically take documents from places like Google Drive, Gmail, or your CRM, and extract important information from them using AI. You can then use this data for various purposes, such as adding it to Google Sheets, QuickBooks, or one of the many other applications available in Make.com.
The possibilities are endless with Make.com and its wide range of applications
Even more applications are available in Make.com to suit your needs
Thank you for watching, and I hope you found this video valuable. If you did, please subscribe to the channel and leave your thoughts in the comment section.