Discover a Comprehensive Software Development Invoice Example for Engineering

Software development invoice example for Engineering

so in this example we're going to work with AI to help a business deal with lots of different expenses and receipts in different formats and to get exactly the data that they need into the tables for further reporting and things like that an admin can use this app to upload invoices from contractors or snap pictures of receipts and since our app works on both mobile and desktop the method of capturing and uploading this data is very flexible data can be extracted directly into different tables where we can use it for whatever the business needs like build building computations on top and Reporting on it highlighting certain records or even using Glides Json tools to send it to another service that they use both of these buttons trigger an on form submit action the invoice button sends data to the invoice table and the receipts button sends data to the receipts table let's take a look at the receipt onsubmit workflow which uses an image to text action when we click on the receipts button it opens up a form and if we click edit here we can edit this workflow firstly on submit we take people to the detail screen for that item and we set the processing column in our table to true this shows our processing image component to indicate to the user that data is being processed next we use the image to text action to extract all of the text from the image the receipt that was uploaded by the user and then we have a prompt that extracts the list of items from the text we have some specifics here in our prompt about how to format the text making sure that it's appropriate for receipts BAS based data we then use another generate text action to get the supplier name and then some text to number actions to extract the total the cache and the change all of this goes into our table as a new row and if there's any issue with the data we give our admins the ability to edit it in an edit form one of the main benefits of this structured data is that we can of course report on it in charts like this so the workflow that you just saw there that was dealing with receipts actually works pretty straightforwardly the image to text action generally captures everything pretty well that's because a receipt is generally a simple top- down structure that's easy to transcribe but the invoices that our contractors send have very very different formats and in fact when we were building this exact template we had very mixed results with the image to text action and column depending on the way that the contract invoices were laid out so depending on your setup you might need to do some trial and error with the instructions in there the prompt but actually if you're finding that it's still not collecting the right information and it's getting a bit confused because the format's a bit all over the place then you might want to try the text to Json AI action Json is the most common way to store and work with data through apis it's a structured data format that isn't limited to a tabular form like spreadsheets instead it uses text to represent data in a hierarchical structure now we're not actually sending this data to an API here instead what we're doing is using AI to turn our unstructured data Into The Very Organized structure of Json and then we're using a non- aai action to kind of query or extract data from that very dense tightly packed structure here we're using the query Json action this is not an AI feature per se but is really benefiting from what the AI has done before because it can easily locate the data that we need in the Json and send it wherever we want so let's break this down a little bit again and go through the workflow step by step firstly we have our show detail screen and the set columns value that triggers the processing screen next we convert the document that's been uploaded to text then we use the text to Json action with very specific instructions we tell it to convert the data into a Json object with the exact fields in this order here we're specifying the data that we're interested in this means that if there's any other data laid out in any other kind of way on the invoice our AI should ignore it and not pull it into structure Json this data then just gets sent to a normal text column in the data editor we can then extract the parts that we need using Glides query Json action firstly we assign the text to Json column as the input to this action then when we open this drop down here we can see that before our column values and the user profile data we have this new area of values that's actually coming from the Json that's been created by the previous action now if you've worked with Json here you know how valuable and timesaving this one feature is you don't have to write queries you can just simply select the data that you need and then send that data wherever you want here you just say where you want that transform data to go and then we repeat that process for all of the other parts of the data that we're interested in and then end the workflow in exactly the same way that we did the previous example if you're more comfortable with the generate text feature instead of Json that's going to work fine for nearly all scenarios but if you do encounter formatting issues with complex documents or you need to send this data to API then Json might be a really good way to go so that's a look at our invoice and expense capture example people can upload images or contracts get structured data immediately in their tables they can view the invoices and receipts in the app and they can edit the data if they need which gives the business a really quick way to capture exactly what they need for reporting and other purposes [Music]