[00:00:00] Speaker 1: So, in this video, we are going to learn how to convert audio files into summarized transcript using AI agent. So, we are going to build an AI agent which will capture the new audio files which we are uploading in our Google Drive folder. Let's say I am uploading the audio files of the lectures or online sessions or of podcast. Then after that, this AI agent will send the same audio file to OpenAI and using OpenAI, it is going to generate the transcription, get the text of the audio file as well. After that, this AI agent will send the transcription to OpenAI or chat.jpg once again and using it, it is going to generate a summary of the same transcript, the same audio file. And at last, we are going to update the summary with the file details and the actual transcription in our Google Spreadsheet. Now, if you want to know how you can create this AI agent all by yourself, for that, let me just take you to my screen. So as you can see, this is my Google Drive folder where I upload the audio recordings of my lecture. Now the AI agent which we have created, this AI agent will capture the audio files from this Google Drive folder and send it to OpenAI. Using OpenAI, first, it is going to transcribe that audio file into text and later on, the same text will be sent to OpenAI once again or to chat.jpg to get the summary of the transcription, the summary of the audio file or the complete lecture. And at last, this AI agent will update the details of the file like file name, file link, the summary it has generated, and the actual transcription in my Google Spreadsheet as well. Now, to build this AI agent and connect all the platforms, we have used Pavli Connect. And to get started with Pavli Connect, we will just go to a browser and search for pabbly.com slash connect. And here you can see that we have reached to Pavli Connect's landing page. Here we have two options, sign in and sign up free. Now, if you're new to Pavli, by clicking on this sign up free button, you can set up your free Pavli Connect account in just two minutes. Or you can just click on the sign in button if you already have a Pavli account. After signing in here, you will reach to all app speech of Pavli. And here we just have to click on access now below Pavli Connect. And as I click on access now, we will reach to Pavli Connect's dashboard. After reaching to this dashboard, we have to create a workflow for our AI agent. For that, just click on this create workflow button and give this workflow a name. Let's say I'll name it as AI agent for audio file summary. You can basically give any other name as per your choice. After naming our workflow, just select the folder of your Pavli Connect account in which you want to set up this workflow and click on create. And as I click on create, here you can see that the workflow will open in front of us. In this workflow, we have two different windows. First one is the trigger window and second one is the action window. So triggers and actions are basically those two concepts, those two principles on which this AI agent, this workflow works on, where the trigger says when this happens and the action says do this. So let's start with the trigger step. Now the idea here is first we have to connect our Google Drive folder with Pavli Connect in the trigger step so that whenever we upload a new audio file in Google Drive, automatically the details of the same file is captured by this AI agent and this AI agent will trigger. So here in trigger step, in choose app, search for Google Drive. Just select it. After selecting Google Drive in trigger event from the drop down, select the trigger event as new file in specific folder. Then just click on connect and select add new connection. Now here we have to connect our Google Drive account with Pavli Connect and to make this connection just click on this sign in with Google button. Then after that here from the pop up window, select your Google Drive account, then just scroll down and click on continue. And as we click on continue, we will see our Google Drive account will get connected with Pavli Connect. After connecting our Google Drive account, the next thing it asks us is to select the folder or subfolder. So in this drop down here, you can see a list of all the folders and subfolders which we have created in our Google Drive account. Out of all of this, the folder in which we are uploading our audio files is named as lecture audio recordings. So from this drop down, we are going to select the folder on which we want to capture the new file. After selecting the folder, when we will click on this save and send test request button, the details of the last file, the most recent file which was uploaded on this folder will be received inside Pavli Connect. So here you can see that in this folder currently we have this file. And if I open this file, you can see that this is an audio file of the lecture. This is a recorded audio of a lecture. But the share permission of this file is private to only me. And with this share permission, this AI agent can't send this file to OpenAI and transcribe it. So what we have to do, we have to change the share permission of this file. And not just of this file, we have to change the share permission of all the audio files which we will be uploading in this folder. So instead of changing the share permission of each and every file one by one, we can change the share permission of this complete folder. Just click on this folder name from the drop down, just click on the share button. As I click on share, you can see that this box will appear here from this restricted drop down, select anyone with the link and then just click on done. And now we have completely changed the share permission of each and every file which is there or which will be uploaded in this Google Drive folder. And now if I open this audio file, you can see the share permission of this audio file now is anyone with the link. After changing the share permission and ensuring that we have at least one file in our Google Drive folder, we just have to click on this save and send test request button. And as I click on save and send test request, you can see that we have received the response. And in this response, we have received the details of the same file. We have the web view link of the same file. If I copy this link and paste it in a new tab, the same file will open in front of us. And after that, we have this web content link. Basically, this is a direct download link of the same file. And we are going to send the file to OpenAI using this link. We'll scroll down further and you can see some other details of the same file. Like here you can see we have the file name and the rest of the details. So in this way, the step one of building this AI agent is completed. Now let's move to step two. And in step two, we want to send the same file to OpenAI and using OpenAI, we are going to generate the transcription of this audio file. For that, just scroll down, come to this action step and here in choose app, we will search for OpenAI. Just select it. After selecting OpenAI in action event from the drop down, select the action event as generate transcript. Then just click on connect and select add new connection. Now here we have to connect our OpenAI account with Pavli Connect. And to make this connection, we have to enter the API key of our OpenAI account. Now you must be thinking how you can get this API key. For that, just log into your OpenAI account and go to the dashboard. From the dashboard, go to this API key section and click on this create new secret key button and give this API key a name. Let's see, I'll name it as Pavli Connect. Then just select the project of your OpenAI account for which you are generating the API key and permission as all. And click on this create secret key button. Now as you click on create secret key, the API key will be generated. Just copy the API key from here, go to Pavli Connect and paste it over here. And after pasting the API key, just click on save. And as you click on save, here you can see that our OpenAI account will get connected with Pavli Connect. After making the connection, here it is asking us for the audio file URL. Now here in the field of audio file URL, we have to basically add the URL, the link of the audio file which we want to transcribe using OpenAI. So we have uploaded the audio file in Google Drive and received the details of the same file inside Pavli Connect in the trigger step. So here from the trigger step responses of Google Drive, we have to add this response of web content link, the download link of the file. And to add the trigger step response to this OpenAI action step, we are going to use a process called mapping. And the process of mapping is very simple. Simply click on this field of audio file URL and from the drop down, from Google Drive responses, select this response of web content link and map it over here. After mapping it, it is asking us for the model. That using which model we want to transcribe this audio file into text. Either we can use Whisper, GPT, Poro, Transcribe or GPT, Poro, Mini, Transcribe. So I will be using the model as Whisper. After selecting the model, just click on this save and send test request button. And as you click on save and send test request, here you can see that within few seconds, we are going to receive some response over here. And the response is received. And this response seems to be a positive response to us. In this response, in front of text, you can see we have received the complete transcription of the audio file. So we have updated or uploaded the audio file of a lecture in our Google Drive folder and the same audio file lecture is received inside Pavli Connect over here and we have the transcription of the same. Now after getting the transcription of the audio file, we want to send the same transcription once again to OpenAI to charge GPT and summarize this as well. For that, just click on this add action step button and then here in choose app, search for OpenAI. Just select it. After selecting OpenAI in action event from the drop down, select the action event as charge GPT. Then just click on connect. And because we have connected our OpenAI account with Pavli Connect in the previous step while transcribing the audio file, we don't have to make a new connection. You can select the existing connection and click on save. And as you click on save, you can see once again our OpenAI account will get connected with Pavli Connect. After making the connection, the first thing it asks is to select the AI model that using which AI model we want to basically send the prompt and summarize this audio file transcription. So in this drop down here, you can see a list of all the models of charge GPT available. Out of all of this, I will be using GPT40mini as the model. Then after that, it is asking us for the prompt. So what is the prompt we want to give to OpenAI? We can enter that prompt over here. Okay, so as you can see, this is the prompt which I'm giving to OpenAI. The prompt says, I have a transcription of a lecture, please summarize the content in a concise, easy to read format. Then extract and list the key points with clear headings. Keep the tone educational and structured suitable for students or professionals. And then here after that in front of here's the transcription, we have to basically add the same transcription or the audio file which we have received from OpenAI from the previous step. So once again, we'll use the same process of mapping, we will just click on this field of prompt in front of here's the transcription where we want to add the transcription. And here from the drop down from OpenAI responses, select the response of the transcription text, and it will be mapped. So we are going to send the transcription with this prompt to OpenAI, to chat GPT. After entering the prompt and mapping the transcription, just click on this save and send test request button. And as we click on save and send test request, you can see that we are going to receive some response within a few seconds. So let's wait for a few seconds over here. And yes, here you can see that we have received the response and in this response in front of message content, we have basically received the complete summary of the same lecture audio over here. So first we have transcribed this audio into text and then using chat GPT, we have received the summary of the same transcription of the same audio as well. Now after getting the summary what we want to do as the last step, we want this AI agent to update this transcription and the summary in our Google spreadsheet. But before doing that, let's delete the existing data which we have in our Google spreadsheet. So we have a clean sheet over here. Now to update these details, add these details in our new row in the Google sheet, we will just scroll down, click on this add action step button. And here in choose app, we will search for Google sheets. Just select it. After selecting Google sheets in action event from the dropdown, select the action event as add new row. Then just click on connect and select add new connection. Now here we have to connect our Google sheets account with public connect. And to make this connection, click on this sign in with Google button. Then after that here from the pop up window, select your Google sheets account. Then just scroll down and click on continue. And as you click on continue, we will see our Google sheets account will get connected with public connect. After making the connection, the first thing it asks us is to select the spreadsheet. So in this dropdown here, you can see a list of all the spreadsheets which we have in our Google sheets account. Out of all of this, I have this spreadsheet named as lecture summary, in which we want to add the details of the transcription, the summary and the audio file which we have uploaded in Google Drive. So from this dropdown, we are going to select that spreadsheet. Now in this spreadsheet, you can see we have only one sheet named as sheet one. That's why in the dropdown of select sheet, we have only one option and we are going to select that. And after that, all the columns of data which we have created in this Google sheet, you can see all of these columns as fields over here. And from the previous step of OpenAI and Google Drive, one by one, we are going to enter these details, map the details into their respective field. Like here in the field of file name, we will just click here. And from Google Drive, we have received the response of the name of the audio file which we have uploaded. So from the Google Drive responses, let's select and map the file name. Then here in file URL from Google Drive itself, we are going to map the web view link of the file. Then here in summary, using OpenAI's chat GPT, we have received the summary of the lecture. So we are going to select and map the response of summary over here. And here in transcription, from OpenAI's generate transcript, we have received the original transcription of the audio file. We are going to select it and map it as well. So in this way, we are going to update all of these details in our Google spreadsheet. After mapping all the details, just click on this save and send test request button. And as you click on save and send test request, you can see that we have received some response. And this response seems to be a positive response to us. This response shows that the data which we have mapped over here, this data, these details are now added as a new row in our Google spreadsheet. So let's check it. We will just go to a Google sheet. And yes, here you can see that in our Google spreadsheet, a new row of data is added over here. And in this data, you can see we have the name of the same audio file of the lecture which we have uploaded in Google Drive. Then we have the link of the same file as well. If I click on this link, the same file will open in a new tab. Now for that, we have the summary of the same file. So this audio file is transcribed using OpenAI using the CI agent. And we have the summary of the transcription of the complete lecture. And we also have the original transcription, the complete original transcription of the audio file too. So in this way, we build the AI agent which will automatically transcribe and summarize audio files for us. So not just these applications, you can connect multiple applications with Pavli Connect. Also let me tell you that you can build AI agents for different processes of business using Pavli Connect and use different or multiple AI tools as well. Also you will get a clone link of the same workflow, the same AI agent in the description box below. By clicking on that link, you can just clone the same workflow into your own Pavli Connect account. And the free plan offered by Pavli Connect using which you will get some free tasks every month into your own Pavli Connect account. You can basically try and test this AI agent for absolutely free. So if this video was helpful to you, don't forget to hit the like button, share it with others. Or if you have any kind of doubts or queries, you can always write us in the comment box below or email us at support at redpavli.com. Also you can post your queries on our forum as well. And if you want to check the pricing of Pavli Connect, you can visit this link. So this is it for today's video. Thank you. Have a great day.