Websites and the step-by-step process of automating data extraction lastly we will cover how to Leverage The Power of AI to streamline workflows with live web data so Julia.
Builds this automation for rajes and make let's create the scenario together we'll first log in to make.com and open our scenario builder then we look for an https module use the https module to.
Fetch the web content in other words to bring the text on that web page into your scenario for this you need to paste the URL in the first field and then set the method to.
Get it will now return the plain HTML content of the website as you can see here now let's convert the HTML to plain text using the tools module in this module all you have to do is map the.
Anthropic we first Define the model in my case clae 3.5 Sonet and the max number of tokens which Define how long the expected output should be in our case we could get a long output from our.
Website so it's better to choose a higher number of tokens like 4,000 here we now add the prompt with instructions to extract the titles and excerpts into a Json format we will show.
The details of our prompt a little later in this video so don't worry about it now now we need to parse the Json output from the AI using the Json module so will just map the response from.
Anthropic into this field no need to worry about creating the data structure now and add the extracted data as row in a Google sheet using the Google Sheets module now let's set up the Google.
Sheet we copy the ID here from the URL bar we paste it in the spreadsheet ID field now we select the Sheet's name which is in that case legal news and set the column range A to Z is enough.
Posts Related:
- Python Part 18 - NumPy Office
- How To Automatically Sync Your Bank Transactions With Excel Free Download
- How To Create A Mobile Inventory Manager App And Sync With Excel And AppSheet
Here after this we take the Json
Module's output and map it into Google sheet once we finish setting up the Google Sheets let's go through the process step by step again first the.HTTP module retrieves the webpage content check their HTML here next the tools module converts the HTML to markdown text making it simpler to process and allows us to stay in the AI.
Token context window now for the Magic in the anthropic module we provided clear instructions with a prompt to the AI our prompt says extract the following information from.
The provided HTML find all the article headlines within H2 tags that have a link for each headline extract the text of the headline as the title.
Field extract the text content of the immediately following paragraph tag as the excerpt field this now returns a collection of title and.
Excerpt so we now need to compile the extracted data into a Json object with keys title and excerpt for each article to actually work with it inside make it is useful for the AI to always.
Provide few shot examples like seen here an example of the desired Json output format this approach we can strongly recommend to Rajesh we should also add to the end.
That the AI should not return anything else than the Json so we avoid errors where a plenty of other information is returned the AI does now the heavy lifting and.
The Json module parses the structured data as we wanted it let's save the data in the Google Sheets module there it adds each title and excerpt as a new row in rajesh's spreadsheet great we are.
Finally ready to run the whole
Scenario we can see that it worked the latest blog post titles and excerpts automatically extracted and added to rajesh's.Spreadsheet so to sum up what we did in this use Case by leveraging the power of AI Julia and rajes turned unstructured web data into structured data that Rajesh could use in his.
Automations this is just one example of how one can use AI to streamline workflows and stay on top of important information congratulations on mastering this use case and we look forward to.
