Lesson 12 - 'By Zapier' Blocks: Fundamentals
Start an AI Automation Business📅 2024-07-31
Transcript
[00:00] welcome to the next lesson here in this module we're going to be looking at the zapier apps these are apps that are native to zapier's platform they're built by zapier and they can be used in a ton of different use cases when it comes to chat gbt outputs just so you're aware of all the apps that are available we got all this list right here and you're going to be able to really use a ton of different stuff when it comes to chbt and these specific apps such as the RSS feed the paths the scheduler or matter and so on so let's go ahead and
[00:32] walk through a couple of these and see good use cases that we will need to use in an AI automation agency looking at the first one here it's going to be scheduled by zapier this is going to allow you to essentially set up a automations that run on specific trigger events so their events is either every day every hour every month every week a lot of context of why you would use a schedule by zapier for chat gbt would be more for Content generation as this typically aligns with social media schedules further more another good
[01:04] context to run something like a scheduler would be for data searching so for example let's say we run a scheduler every Sunday and then you're looking for specific messages found within a slack Channel specific inputs found within a Google Excel and so on now let's go ahead and pair the schedule by zap year with paths in order to create a really cool social media schedule in order to do so we would just add another step here we'll make it pass and essentially paths are think of them like multiple filters
[01:35] so depending on whether it passes through the quote unquote filter as you see here it will proceed with the automation so in the context of creating a social media schedule let's just say that we go ahead and set this scheduler to every single day it's triggered hit continue here time of day we'll just say midnight continue we're going to test this trigger and as you see here the data that's provided is that it gives us pretty day of the week so let's say we run a social media schedule but we only
[02:06] want to post on Monday Wednesday Friday we need to do is come to our path a here and we would say essentially pretty day of the week exactly matches Monday therefore that when Monday and this trigger incurs it will continue and it'll go down this path so let's go ahead and say that we want to keep whatever content is being generated on this path the same we're going to say and or stay here we're going to say pretty dat of the week exactly matches uh Tuesday so essentially every time it's
[02:38] Tuesday or Monday it will go down this flow let's say for Friday we have a different piece of content and we still want it to get triggered we'll do the same thing here exactly matches Friday and as you see here it won't work because it's not Friday today but what's great and what we set up here essentially is that every Monday Tuesday Friday this flow will run and that could be your social media content schedule so so another buy zappier tool we can use here is RSS now RSS is going to be pretty fundamental when you're dealing with a client that either has a podcast
[03:09] or some type of content feed that they've linked to an RSS feed using an RSS feed all you really need to do is simply is either do new item in feed or new item in multiple feeds depending on the context you continue and then you would just input the business's feed URL here and then everything following that we can ask and Proctor with chat gbt now as a real quick rule of thumb and really cool thing you can do with Chad gbt is let's say you run an RSS feed and for that specific feed you want to look for a very specific type of content that incurs within that feed we're going to
[03:39] go ahead and create a conversation here within this conversation we're going to Simply put base on the new item in the feed you do parentheses and then you would just input the data point right here going to enter generate yes we do semicolons yes or semicolon no on whether it has to do with cars as an example whatever the business context you may be in you would put that there
[04:11] and essentially what we can do here is we can use 3.5 because we'll be able to understand just the sediments and the the data within that feed hit continue here hit and whether it's yes or no will depend on whether the flow continues if it's yes then we all we would need to do essentially is continue the flow but to ensure that it doesn't continue no matter what we add a filter block here and we would say based off the assistant response message so let's go just test this action we would want it to say exactly matches no now in this context that did come out a little bit longer because we didn't provide an RSS feed
[04:42] but in the context we did it essentially would just output either yes or no allowing you to do more of data analysis in a potential flow now a really powerful block you can also use in zapier is a code block now I see using Code blocks more in other use cases when it has to do with other apps rarely you ever find yourself really needing to use a code Block in the context of you know ensuring an output for Chad gbt most likely you will be putting outputs into the uh event here so essentially what I mean by that is that within the code
[05:13] itself you would put an output from J gbt rarely do you ever find yourself needing to Proctor something this is more for advanced AI automations that require you to kind of manipulate the outputs of uh third party apps I'll show you a really good use case for this in our AI article generator video now another byy zapier app that you can start utilizing within your flows is going to be a delay now you might be asking yourself why would I use a delay in the context of AI automation do I want it to be fast well there are some use cases for that so think of it this way let's say you had an AI
[05:45] autogenerated email for a specific customer when a customer sends that email you shouldn't be responding to it right away using a delay block we can mimic human behavior here maybe add an hour delay 2H hour delay 30 minutes 50 minutes and so on this is going to be allow you to take AI automation flows and mimic more human behavior which can be useful in certain use cases depending on the AI automation another cool app provided by zapier is the storage app I really like using this one this one essentially allows you to take data
[06:15] that's been generated from one flow and then pass it to another one so what do I mean by that essentially what you can do is let's say there is a per piece of information an ID a specific output from jbt that needs to be taken and given to another flow that incurs later on in the process but they're two separate zaps how do I do that essentially what you do is this you will go ahead and say uh set value continue continue here you'll give a key this is we're going to call upon it we're just going to call it uh gbt 4
[06:46] it's this could be anything so we do gbt 4D the value that you want to save for the other flow so go ahead and just save the resistant response message Continue test that action and then we're going to go ahead and save that so when you create the other automation so in the other zap how you call upon that is simply doing another storage block to say get value continue continue and then remember we named our key gbt 4D hit continue here and then this should show up as what we had
[07:18] earlier perfect it does and that's how I grabbed the data from one zap your flow and passed it to another so another cool feature zapier has is a web parser so this is going to allow you to grab grab information from the internet that may be relevant to what you're doing so an example of this could be possibly that within your automation you get the link for an article that was just created or there was a link found within an email you're going to want to go ahead and use the parse web page you would provide the URL there that you find and essentially
[07:49] it would parse it out in HTML mark down or plain text and from there on that output you'll be able to use gbt to analyze it and give a specific formatted output of whatever you wish now let's get to the very famous one here which is probably going to be the one you use a lot within your a automations which is going to be the formatter by zapier now the formatter allows you to do a couple of things such as date time numbers text and utilities essentially whatever you plan on manipulating is what you're going to be choosing for the event most
[08:19] of the times when dealing with Chad gbt and a automations it's going to be text next we have the option to transform so this is going to be all the different things we can do with that underlying output so knowing so let's go ahead and run through some use cases we're going to go ahead and structure this prompt again we're going to go ahead and say generate two social media captions for an ice cream company and we can go ahead and keep this um at gbt 3.5 as it's not too important let's go and just see what
[08:49] these outputs would look like all right perfect so as you see here we got our two outputs we got the first caption and the second caption uh let's go ahead and use the formatter but one thing I want to show you y'all real real quick is that if you're having issues with the form matter and you're really trying to do something but you can't really understand or know how to do it there's one new thing that CH or zier had added which essentially allows you to format the data using AI this isn't perfect but it can help in some use cases so essentially what you would do is You' come to the assistant response message you would hit format data here and
[09:19] essentially it asks in layman terms what you want to change so in this context since I see those quotation marks I could ask it um remove the quotation marks from text so then it'll try to interpret that and try to make its own formatter block with zapier so let's go and see if it was correct so it was correct in this use case if I hit continue here and test this action it would remove the quotation marks now this feature isn't perfect but it could lead you in the right direction so I might as well show you it so in order to
[09:51] do that without the underlying use case of the AI we're going to go ahead and just do replace so essentially what the replace function uh does which can be very useful in multiple use cases is that it'll find the specific character that you're looking to replace here such as the quotation mark and then you would be able to replace it with either empty space or an actual space depending on your context so hit continue and test this action also note in order to call upon certain functions you're going to have to use their phrasing so as you see here this is how you call upon an actual
[10:21] space a tab a new line and return that is one use case to use formatter there is another one here called split so what I like to do with split is if there is a long data output sometimes you want to have it so the data is in just one block to call upon and what I mean by that is essentially notice how this assistant response message is both captions in one data piece so in order to make it so that it is two captions in two separate data pieces what we're going to do here is we're going to go ahead and get the Imp output again and this may take a couple tries here but let's go ahead and
[10:52] try the first function here as I showed earlier which is going to be new line so we're going to go ahead and format it like it likes it we do set index here and what's very important here is I'm going to do all separate Fields this is going to allow it to be separate data points we're going to test this action and there you go as you see we have two separate data points and what do I mean by that essentially if I come over here go and delete that we got the format text we got output item one shift shift outpoint item two if I didn't format it like that and I just did the
[11:22] original chat gbt message it would have kept the two captions together but in some context you need to separate data so it isn't just one block now there are other use cases that I've used in the past such as the capitalized one that's really cool essentially all that really does is that it'll take the underl text and then capitalize every single word there we go every single one of these words have been capitalized there are other really cool functions such as extracting an email address from a block of text extracting a number extracting a
[11:54] phone number looking for length of a text getting the word count for the text if that is important in some use cases from what I've shown this essentially allows you to do a lot of functions when manipulating the underlying data found within the chat gbt output now one thing I'll point out is that if you want to do something very very complex with that underlying output and you can't seem to Proctor it from chat gbt that will typically require you to use a code block but what's great about code blocks is it does allow you to use AI as well
[12:26] as you see here you can generate with AI sometimes it works sometimes it doesn't if it doesn't work then what I typically like to do is either use chat gbt and stack Overflow or just use chat gbt 4 in general and really proct your answers for whatever you're trying to achieve for reference currently all zapu uses is either python or Javas script okay I hope you learned something about the zap your blogs and why they're so fundamental for specific use cases when it comes to AI automation there was one main one that we did not mention and that was web hooks that's
[12:57] because in the next lesson here we have a whole video dedicated to web hooks as it's more complex but honestly it's one of the more valuable ones to learn about because of the fact that it's going to allow us to do custom actions within zapier's interface so let's go ahead and learn how we do API calls and really understand more complexities when it comes to AI automations