Automatically Research Anyone with ChatGPT and Perplexity

LARRUZ shows how to set up an automated research workflow using Make, Notion, ChatGPT, and Perplexity to pull LinkedIn profiles and generate detailed reports. The setup costs about $0.04 to $0.07 per person researched and takes minutes to configure with a prebuilt blueprint.
Chapters
Today I have a very special video for you that is going to allow you to use the power of AI and the data that's on the Internet to research various people. Now this can be useful in several scenarios. For example, you might be meeting with somebody you haven't met yet and gathering some extra info on them before the meeting could be helpful. If you think about it, powerful people like politicians have assistants that whisper into their ear over the shoulder that prepare details about the people who they're meeting. Common people don't have advantages like this, but with the power of AI and automation via make here you can start implementing research workflows like this into your career or everyday life. But instead of me telling you about this, let me just hop on over here and show you what you'll be able to do at the end of this video. It's all starts in this notion database. And yes, you could customize this to other software, but this is what we use. So instead of this notion if I just create a brand new row, all I need to do is inside of this profile link field I will share a LinkedIn URL, hit Enter and switch the research status to initialized. As soon as this is initialized, the automation in the backend scans every 15 minutes and researches the person from the sequence LinkedIn profile. In this case I will manually start it and press play here. And if you're non technical, no worries, I will share these blueprints with you and show you every step of the way on how to set this up on your own machine without any other knowledge required. And with a little bit of patience you will see that the scenario here was initialized and now it should be going through these various steps one by one. If I head on back over to my notion database you will see that this automatically switched to in progress. And already I have the name in here because this thing is going to work. And searching not just LinkedIn but also using perplexity search in multiple ways to gather data on myself in this case. And there you go. In under one minute this has switched to done and now I can open up this entry and look at that. All this info on me that doesn't just come from Google, it also uses perplexity to gather some of this. And all of a sudden I have a blog post that has been written automatically on any person that you can find on LinkedIn. Now this is pretty amazing and I found services online that sell something similar to this, but it's not customizable and it will cost you money. All you need for this thing right Here is a Make subscription and your OpenAI and perplexity API key. Now before we get into the rest of the video, which is going to be all about showing you the step by step of how to import this and activate this for yourself, spoiler alert. It's very simple and non technical. Me and the team prepared this entire blueprint so you can just copy paste this into your own MAKE account and get going in meantime. But before we do that, I just want to point out that this video is sponsored by make. But I was so excited when they reached out because we were already using their automations and teaching how to automate with make because the user interface is so intuitive and their free trials allow you to get started and run automations like this for free. And then later on you can make a decision if you want to pay for the subscription to keep this going. One more note, as you might see, I'm not in the studio, I'm actually in the north of Brazil, concretely Jerichoara on an extensive kite surfing workation. And I'm sharing this because I think it's so relevant because the goal with all of these technologies is to free you up. They should open up your calendar and give you back the most valuable asset your time. So you can do whatever you desire while these automations go to work. And instead of sitting there and googling and using perplexity and researching on LinkedIn and an automation like this can do it all for you. And now I'll show you how to build this yourself. So first things first, you need to sign up for a MAKE account. I personally did this by just logging in with my Google account. You can start up with a free trial and give this a go. And all you need to do is go over here to scenarios, say create new scenario. And here's the wonderful part. MAKE actually supports blueprints that support save all of the presets. A lot of other automation softwares don't do this. So this is one of the reasons why I really love make. I can give you a complex preset like this and you can just import it and you'll very simply do that by going to the link in the description and downloading this file. It's called Web Researcher Blueprint JSON. And all you need to do in here is say import blueprint. And now you might have already guessed it. You just navigate to the file that you just downloaded for free from the description you say save and voila, the entire thing is imported. Now this is not the end of your journey. You do need to activate a few connections for this to work with Notion, chatgpt and Perplexity. So let me guide you through that step by step so you can make sure this works as fluently on your machine as you saw it working on mine in the beginning of this video. Okay, so the next thing you want to do is you want to go to the next link in the description. And I just prepared this sample database in Notion, which you can duplicate to your account. If you don't have Notion, I would recommend you give it a shot. You can also create a free account and just duplicate this to your own account by just saying duplicate and then you just add it to your private pages like so. And now in your private pages you have this brand new DB Researcher template database that will allow you to follow along this tutorial and then you could use it to research wherever you like. You could absolutely customize this to use something like Google Sheets. But that goes beyond the scope of the tutorial. This automation has been set up to work with this page. All right, great. So now you have the blueprint, you have the Notion database. Now all that is left is linking everything together and making sure that the automation also Splits speaks to ChatGPT and Perplexity AI. Luckily, all of those are not very hard. Let's just begin with Notion. You will want to start off on this link Notion. So profile integrations again you can find in the description and here you will simply say new integration. We will call this LinkedIn researcher tutorial 1. I will link it to my Notion workspace. In your case, if you have a new account, you should only have one here I will keep this as internal and say save Save. Simple as that. And then if I head on over to Notion, I can click the three dots in the top right here. Here in the bottom I can say connect to. And in this link you should be able to find this LinkedIn researcher tutorial 1. Yes, this is it right here. I say confirm and voila, your connection has been set up. Now we can use this connection inside of make to easily set this up for yourself. So what I need to do now is I need to click every node here that has the Notion logo. You see there's 1, 2, 3, 3, 4, 5, 6 of them. So let's begin with the first one. And on the first one I will need to add this brand new connection here on the connection type I will say Notion Internal. And right here it will ask for my internal integration token. Okay, and this is very simple. I Just head back to this page where I created the connection, say show and this is the integration token. Okay, so I just copied this into here. I say paste, save and now my make account is connected with the Notion account. So in every one of these I can just go in here and say my notion internal connection. And then I can link this database which I'll just quickly rename so we have an easier time doing this. For the tutorial DB Researcher 1 here on the database ID, I say search. If I just say DB Researcher 1 in here and okay, it will automatically find this database and link it up. I say okay, and that's it. Now okay, fair enough. Those were a few steps. The good news is for consecutive ones it's way easier. Just say okay. Here. For the next step we do the same thing. We just pick the connection that we just created. In your case you should only have one, right? Here we search for the database, input the name, say ok and it will automatically find it. For next one I'll do the same thing. The internal connection search for the database, input the name, say ok and done. I think you see the point. I'll just repeat the same thing for the last three nodes here at the bottom. 1, 2, 3. Let's just fast forward here. Okay, that concludes the first part and that's actually a bulk of the work because all that is left now is for you to set up the API key connections to both the OpenAI API and the Perplexity API. These two connections that we will set up now allow for the Automation to use GPT4O and the Perplexity web search. Let's begin with the OpenAI connection. All you need to do for that is head ON over to platform.OpenAI.com API keys. As with all the other links, you can find them in the description of this video. And here you will create a new secret key. Let's call this one make tutorial. And immediately it will create this key that you can now copy and back inside of the make scenario. All you need to do is go to one of these OpenAI steps and you will add a new connection through this button. I will just rename the connection to make tutorial as I have multiple here and simply paste the API key in here and say save. Now as you can see, everything else happens fully automatically. This template comes pre filled with the full prompt that you can customize but you don't have to. I just click ok and now I have to repeat the same step one more time on this ChatGPT step. Simply pick the Make Tutorial connection that we Already established, hit OK and we're done. This is connected to the OpenAI API which allows you to pay per usage. Next up, Perplexity. This is equally simple. Once inside your Perplexity account, you just need to go to Perplexity AI Settings API. And here I can generate a brand new API key which I will now copy. And back in my scenario I click on the first Perplexity step up here I say add again. I will rename this one to Make Tutorial and I paste the API key in here, say save and everything will automatically readjust. Amazing. So now I hit okay and all that is left is readjusting the two remaining Perplexity steps so they also are connected. One of them is over here, make tutorial, hit okay. And the last one is over here Perplexity, make tutorial and hit okay. And that's it. We set up the entire connection. Now the only thing that I will point out here is that you need to take care of having some sort of box billing or credits inside of your accounts. So in the case of OpenAI, if you head on over to the settings and billing, make sure you have some credits loaded here. As you do pay per usage. Every request with OpenAI will cost you a few cents. And the same thing goes for Perplexity. You need a few dollars in here for this to work. And in case you're wondering how much this costs outside of the make subscription, every time this automation runs it will cost you somewhere from 4 to 7 cents, which I dare say is a pretty incredible price for something that would probably take you about 20 to 30 minutes if you do it manually. Okay, so now we're finally ready to use this thing. We set up the full scenario with all the connections, we set up our Notion database and now we're ready to use it. I'll do it again on top of my profile. So I'll Simply copy my LinkedIn profile URL, head on over to my notion, add a new column and I paste the URL in here. And once I set the research status to initialized, this automation would begin if scheduling was on. So in order to turn scheduling on, you would just need to flip the switch to on. And now every 15 minutes it will check the database and see if you switched anything to initialized, which will then make it run. You can already see this is in progress. It checked right away, it saw that this line was initialized and now it's doing the research. If we just wait for a few seconds here this automation is really fast. You will see that soon it will auto populate the name column and then start inserting all the research it is doing right now. So this takes a few seconds, but this in progress indicator tells you that it's already working on it. And there you go, you can see it already appeared. And if we look in here, it will be adding these different paragraphs with one by one by itself. Look at that. It's just writing for us. Isn't this amazing? Once it's complete, this will switch to done. Now while it does that, I want to show you the second alternative which is simply pressing this run button. And that just makes it run instantly if you need that. But the most convenient way is to turn on scheduling. You could even set this to scan every single minute. And there you go. The research has been completed. So if I look in here it will have research on my personal background, my education and all the other info I managed to pull in both LinkedIn and the perplexity searches. One more note is that if you want to improve this formatting, you can cut all of this out and with Command Shift V on a Mac, repaste it and then you have the nice markdown formatting. This is a thing in notion where we just didn't find a way to automatically format it like this. My last note here will be that you can actually go into these different Perplexity and chatgpt steps and customize the prompts. So if you're maybe a recruiter and you're looking to hire for a specific position, well, you could put into the prompt to pre screen them for you to see if their CV actually lines up. You could do everything that you can do inside of GPT4 or like project prompting it to score the candidate on a scale of 1 to 10 on different skill sets. Or maybe you're not a recruiter, maybe you're doing something completely different and you're just interested in good conversation starters with this specific person. Well, you could add another bullet point to the prompt to give you ideal conversation starters that will resonate with this person and so much more you can imagine. The possibilities here are endless. And there you go. That's a full automation on how to set this up for yourself. I personally think this can be really powerful, especially if you're actively networking meeting a lot of people. Heck, you could use this at conferences before you talk to a person. You could just put their LinkedIn in here. It will do all the research for you. You can quickly ch it on your phone no need for your laptop. You don't even need to click anything, you just need to find their LinkedIn and copy paste it into the notion. And if you think even beyond that, you could extend this with other data sources. You could add a simple step in here that will upload their CV and ChatGPT will consider that while writing the final step. But that goes into customization territory. This tutorial was really all about showing you how to copy the blueprint. Set it up for yourself and start saving time today. I genuinely hope this was valuable. Thank you again to MAKE for partnering with us on this video. And make sure to check out the very first link in the description to sign up for MAKE today and give this a shot yourself. All right, see you soon.










