How to receive a daily email with the summary of your favorite podcast made by artificial intelligence

Let’s explain to you how to receive summaries of your favorite podcasts by email. These summaries will be made by artificial intelligenceafter you have received the audio file and processed it. This is not something you can do in a specific application, so we will have to configure a workflow of several steps.

Specifically, we are going to use the service Makebecause its free version is enough to do it. In addition, we are going to use artificial intelligence Geminisince it is very easy to obtain a free API with which to use AI on this website.

At Make.com we will have to log in with the Gemini modules and especially Gmail. It is important to highlight this because in the end you will be giving a third party access to your email, which could compromise your privacy, although this tool is reputable enough to think that it complies and that you should be safe.

First you need the RSS of the podcast

In order to carry out this automation, the first thing is get the RSS address of the podcast that you want to analyze. It is the page with the code where the title, description and the address of the mp3 file of each chapter appear, which in the end is the most important thing.

If you don’t know where to get this address, you can use the Get RSS Feed websitewhich you can get at getrssfeed.com. On this website, you just have to enter the URL to Apple Podcast, iVoox, Spotify or the podcast website, and then it will search there for the RSS address. When it does, click on the button RSS Copy and that’s it, you will have the address copied.

It will also be important that Find out what time each chapter is published. At the end of this project, the workflow will have to be scheduled to activate at a certain time each day, and it should be at a time when the chapter has already been published.

Configure the Gemini API Key

The AI ​​we are going to use for this project is Gemini. And to link and use this AI, you need get a google api. For that, go to the website aistudio.google.com and sign in with your Google account. When you do, in the bar on the left at the bottom click on Get API Key.

Now you have to click on the option Create API key that appears at the top of the screen you have created. This will open a window where you have to create the project for which you are going to use it in order to identify it, for example Gmail Gemini. When you create the project, you can now create the API.

When you have created the API, you will see that it appears in the list of API keys. You just have to click on the left, below where it says Clueand a window with the API will open, starting with “AIza–“.

When you later create the first Gemini modulethen on the login screen you will have to write the Gemini API Key that you generated at the beginning of this tutorial. We are not going to repeat this later because it is something that we will assume has already been done after previous projects such as create a Telegram bot that summarizes your emails.

Now we create the workflow

Now let’s go to Make.com and create a new scenario or workflow. We start by clicking on the initial + for the first step, for which you have to choose the option Watch RSS feed items. It is used to display the elements of an RSS address.

In this first module, what you have to do is write the RSS URL that we have searched for in the previous step, and put it in the URL field. Then, in Maximum number of returned items set 1, so that every time this scenario is run it only pulls the latest post.

Once you have this first step configured, I recommend you click on Run eleven to run the project with just this step. Thus, the content will be loaded, and it will appear in the autocompleted suggestions in the following steps.

Now let’s go add module HTTP with the action Download a filewhich will be used to extract and download the audio file of the podcast chapter. When you click on URL, a window will appear with autocompletion suggestions, and here you will have to write the address of the mp3 file of the audio. If necessary, you can search for it in the RSS code. In our case, it is in Enclosures > URL.

This is a somewhat complex step. A little tip I can give you if you’re not clear is to ask an AI where the audio file is to download, and attach the RSS web address.

After the HTTP module, we are going to add a Gemini module. Specifically, we must add module Gemini Upload to filewhich is used to upload a file to artificial intelligence. Here in the field File you have to choose the option HTTP – Download a file. This means that the file that we are going to upload will be the one that we have downloaded with the previous module.

Now, After the Gemini module we are going to put another Gemini module. We are going to use the module Generate a response to generate a response. Whom Role put User, and in the Partsin it Item 1 set the message type as text, and choose the option Source to add the source text. and in Item 2 put Filefollowed by audio/mpeg in Mime Typeand in File URI look for the URI. All these elements will have to come from the Upload to file module.

Further down, in the field of System Instructionsyou have to open the section Prompts and write a prompt in the field of Prompt 1. This is the prompt with which we will ask Gemini to generate its summary of the podcast audio file you processed. You can write something like this:

Listen to this podcast episode and generate: 1) The name of the Podcast, the chapter number and the chapter title, 2) A summary in 3-5 paragraphs of the podcast content, 3) The 5 key points. Respond in Spanish.

Here, you can customize the prompt whatever you want. For example, you can change the number of paragraphs in which you want the summary to be made, and also change or remove the one that makes key points. Ultimately, you can change it however you want depending on the type of summary you want to get from the podcast. You are in control.

Now we come to the last step. Let’s add a module Send an email of Gmail. In it, as sender put your email address, and in Subject Enter the title that the emails will have. In body type you better add Collection of contentsand where it says Content 1in the field text add the option Result of the module Generate a response.

With this, what you are going to do is that the email sent to you has the result of the prompt from the previous step, which is precisely the summary that we have asked it to do.

Now we have to test and program

Now we have to test if our entire chain of events works. For that, press the button Run elevenand if it doesn’t seem to work, click the down arrow on the right and click run scenario choose the one that appears in green from the first time you tried it when you first made it. So, if it doesn’t work because it doesn’t find anything new, we will launch it again as if it were the first time, with the last chapter we got in the first test.

If all goes well, you should see the progress of your scenario, and when it reaches the end everything should appear green and you should receive the email with the summary. If something has appeared wrong, then the module where the error is will appear in red along with a message telling you what has gone wrong.

When you see that everything is working well, Click on the button to configure how often it runs this workflow. By default it will put you Every 15 minutesevery 15 minutes. Here, it is best to click on this indicator, in the field Run scenario choose Everyday to launch it once a day, and below choose the exact time at which you want it to run automatically. It is also important to click on the save button that appears with the hard drive icon.

In Xataka Basics | How to create a Telegram bot that sends you a summary made by Gemini of each email you receive in Gmail and other emails