Brief intro:

At Gather Real Estate, there is a big need for evaluating a lot of property assets. On any given project, we could have someone selling 10 homes, or 100+ homes and sometimes their located together, other times they could be located all over the city. So we needed a way to be able to take all of the property addresses on a spreadsheet we've received from a client, and then fill in any of the lacking information we didn't have.
For example, if we were evaluating a property, we need to know things like total square footage, how old the home is, what school district it's in, how many bedrooms and bathrooms it has , etc.
Sometimes the client is able to provide this, but often they don't have that information readily available.
Doing this manually for 10 homes takes about an hour of work if you're just getting the basic info and you're moving at a decent pace.
This is why we decided to automate the task and ultimately invest in using an API data source to provide us with the information we need.  

My process: 

I started by evaluating several different companies that offer tools that could do the job with automation (it's always good to price compare) as well as looking at what it would take to purely automate the entire process off of pure web scraping such as going to Zillow and "manually" having a bot perform the same search a person would do to find the property.
Ultimately we opted to look for  an API that would allow us access to the data we needed. This would create a much more stable application that runs smoothly into the future without much upkeep.
After evaluating  3 different data sources, I decided that RentCast.io was the cheapest platform that provided what we needed and allowed us to do testing on their API without the need to purchase a license. So I would be able to build out an MVP of the automation without paying for the API. 

I started working with Windows Power Automate Desktop to take the address data and then run it through the RentCast API call process.
Initially, I wanted to test run with one property at a time but, eventually, I created a custom Ui that would allow for single properties, multiple property addresses copy/pasted into a text box, and uploading a formatted spreadsheet into the program with addresses on it. 

I built out four different use cases for each address. Use case one is for properties that we needed basic info for such as square footage or when the property was built. Use case two is for properties that we needed sale comp values, and use case three is for properties that we needed rental comp values. Use case four is any combination of task 1-3. 

Final Outcome:

The final product has increased our analysis capability by a huge margin. Simply put. the ability to quickly gather all needed property details and multiple forms of pricing on as many houses as we need for a given portfolio gives us the speed of being able to determine price values, realistic sales prices, and potential offer prices from investors.