what's going on guys this is Oliver formerly from response Ai and now from trome. com and get sea and in this video I'm going to be talking about how in every SAS product I've ever built that was profitable I've leveraged other people's code to achieve the end result for the user right and I've made hundreds of thousands of dollars doing this and in this video I'll explain how I did that and what you guys can do and what you can build with other people's code also known as apis and while you might be over complicating just getting some customers via SAS right so the most valuable skill in modern entrepreneurship and SAS specifically is not code or design or sales it's just recognizing patterns right people want things and you can give them these things at scale using SAS right now the ability to see gaps and a sort of Arbitrage between what an API or other people's code enables you to and what customers need this is the art of API Arbitrage and it's the closest thing we have to Magic in 2025 right so the Lego philosophies so explain what an API is um obviously people who already know that's fine it'll be a nice way of you know sort of uh understanding it in a different way and people who do not know then this is the explanation right so an API imagine you had a party where everyone brings half-built Lego sets right let's assume that you're building a boat now your contribution is a boat that is missing its sales and someone else has a pirate flag but no ship Etc right so an API is this agreement to trade pieces to make an end result or an end product that works so you pay a small fee for their Pirate Flag you attach it to your boat and suddenly your creation is a complete pirate boat and it's worth twice as much right it's a complete functioning pirate boat this isn't just sort of collaboration it's it's multiplying the value and as long as you can pay people or some things are free you can get whatever pieces you want to build the most beautiful boat in the world right now the magic lies in recognizing that no one cares guys how you built the sales in the same sense no one cares how you built your software customers only care listen to this customers only care that the boat floats and looks compelling and in this case customers only care that your SAS delivers the result that they want right but developers often gatekeep this right and they insist that you have to knit the sales yourself or what if the Young it's an organic what stit this is called the tyranny of craftsmanship and it conflates technical Purity with business value in other words the only thing that matters is your how your product is valuable to the customer right not how it's built Mark Anderson said in 2024 the best startups today are curators not creators so we are looking to curate tools that we can build with and then build on top of that and then charge the customer right now the epidemic of over complication so there's a conspiracy in Tech right so Engineers inflate complexity to justify their expertise they'll warn you about things like rate limits or authentication or edge cases or security and all of these are valid concerns but they are not valid or or relevant until you've you know past a certain threshold of say 100 customers right the goal guys for building a sass is not to build a faultless system the goal is to build a system that survives long enough to matter in the market and provide value to some customers because if you provide value to some you can provide value to many right now when it comes to sort of using apis this is where the chat gbt wrapper comes from right so it's this phenomenon where every SAS tool these days is just called a GPT rapper right all that means is we took boring tools and we made them extraordinary by adding open ai's API or in other words the chat gbt API now remember an API is just you paying to access code that you couldn't build yourself and I really I personally do not want to try to rebuild chbt to just save a few pennies right so I'm going to I'm going to access their code and use it in my project to make much more money and help much more people five years ago for example customer support Bots followed these rigid scripts um are you a customer yes what is your email X email um do you have a receipt yes right now with ch hbt and and and the apis they can mimic human interaction you ever speak to a chat bot and it's like hey Steve you know really nice to see you here again I love that you recently purchased this it's crazy right this is because the simple app which was customer support just a yes no you know sort of simple flow it's accessing apis to make it more powerful right so we go back to that example it's completing the Lego boat it's taking the half finished boat the pirate flag and it's creating a pirate ship that we can sell for much more money it provides much more value another example right so this this is how you can commoditize the expertise of you know researchers and things like that and make a lot of money so an example is you scrape LinkedIn jobs from something like appify which is just an API Marketplace I'll explain that later you analyze the sentiment of the jobs with open AI like you know um for example the salary um what expertise is required what uh experience is required what degree or what college you know um thing is is required and then you generate a PDF Report with PDF shift right so that is three apis guys that are acting in tandem so you scrape the jobs with the LinkedIn API you analyze the sentiment with open AI API and then you generate the report that's three apis that you are using to create the end result which is the PDF report right the total cost per user if you add it all up would be for example 0. 013 cents right then you just charge 0.
05 and you've got a pretty significant margin right so you scale to one you know 10,000 users and you're netting hundreds of dollars a day for moving data between pipes you don't really touch much of this you just take the LinkedIn jobs you scrape them you analyze them you generate the report and then it's done that's just like a simple example guys the real problem Prof it lies in asymmetrical pricing guys right so API often charge per call or per request right while customers pay for the outcome of your software so this is where it's important my last SAS uh response AI charged around $99 a month for 10,000 AI video credits right now knowing that every 10,000 credits costed us x amount in server costs and the apis I could charge for the margin and that's where you build the product you know profitable tool guys right so users rarely Max their quotas but the perception of abundance like got 10,000 it Justified the price right so the hidden cost of Independence so the most profitable Founders I know they treat apis and accessing other people's code like a tax on progress right it's a small fee guys to avoid having to reinvent the wheel every time you build a new software so when I built a tool to scrape and analy Iz Twitter um for sentiment on Apple stock as a just a hobby project I could have spent months reverse engineering X's you know anti-bot measures and how to scrape Twitter and stuff like that launching my own servers guys like ah man instead I just paid one cent per result via rapid API or um appify whatever and focused on what users actually wanted which is the real time alert about Apple sentiment like people aren't happy with this people are happy with that the earnings report came back all the stuff in a way that they could understand it and I sent them an email every single day explaining it right and that is how you build a simple app I didn't have to learn how to scrape Twitter I didn't have to learn how to analyze stuff with AI I just used Twitter's um apify API and then I just sent it to open AI to analyze right and that's simple the simple sort of pipeline now the 1% rule guys is if an API solves even 1% of your problem you use it right so your job is to stitch the remaining 99% into a narrative that customers will pay for so in other words when you Stitch this pipeline together it has to create the end result the example is the um PDF report on Apple stock your sash should create a result at the end that that users like right now the less apis you use the better but you don't have to kill yourself over it right if you can't build something it is okay to use an API so this is the apathy store guys so I just wanted to talk a bit about this like really quickly right so as you can see um there' be a link in the description to this this is just a few of um the scrapers from apathy right and you see how we can stitch these together with other apis to create what is effectively a full scale tool right so what you would do is let's say we want to create a we scraper that analyzes websites and allows you to ask questions about websites so for example I scrape apple. com and then I can ask what is the price of its most recent product or what's some recent news about apple right so just a double just a double confirm we've got a tool that scrapes a website and then creates like a knowledge base for you to ask questions right simple right you are just going to use the web scraper which is the first result there you going to craw the website for the data and then you were just going to send the data to open AI to analyze right The Prompt would be something like analyze this website um and note all of the important parts of it and tell me what their recent price is for their most recent you know launch product this is how you've just immediately built a SAS without ever really touching code at all you're you're creating the pipeline with a bit of code like a bit of a front end and stuff which I talk about in my last last video but you're hardly building the server to scrape a we you know a website you're hardly building the AI large language model and spending three years doing that to to analyze the tool H to analyze the data you are just sending the web scraper data to open Ai and then you've got yourself a very very simple SAS right now if I think of just like 10 ideas out of the blue that take realistically they take zero effort but it's all about the execution right so you can just ask AI for different examples so let's say I want to you know create an email finder what I would do is I would take clear bits API about scraping social media data and then I would layer hunter. I's API finding emails on top of that and you can just sell it as a leaden tool that the tool could be you know enter a thousand LinkedIn URLs and you get all of the emails of the users right and then you charge $200 a month for it guys like I know it's I know it's kind of just like these are vague examples but again an AI invoice generator right so you take stripe payments because they have an API and you send it to chat gbt to manage and create an invoice right social media right use open AI to suggest posts about a topic and then use Buffer's API to schedule them right so this is how again how apis are talking to each other and an example is you could create a tool that scrapes LinkedIn for profiles and then adds them to HubSpot API so that could be based off of a company that could be based off of a job posting anything like that then something like a no code chatbot so embed chat gbt as a chatbot into a website and then you just add the context for the support from the users business so like you know you would feed chat gbt um there's hundreds of different tutorials guys on how to make like a website chatbot with open AI it's one of the first things that people ever built right so you've got like chat base which is an amazing tool guys that like basically you know does this it it uses the the person's business details as like training data and then you can ask it questions about the the you know the business so you could go to a chiropractor website and say like oh what time can I come in on Tuesday and it'll just say 9:00 a.
m. because that's the data it's trained on right again these are just random ideas it took zero effort just asked AI for these like oh build me API based tools and give me some examples right now this is important guys the API Gold Rush and its traps right so platforms like appify and Rapid API which are just these giant marketplaces for apis um they're they're amazing and they've built millionaires right and half of my SAS tools I have about six software tools right um two of which are making thousands of dollars a month I use apify in both of these right in both of the profitable ones that are actually scaling up and doing well right now in total have maybe you know 100 customers between the two right chaining apis does compound right guys so a one cent call in 10 steps means that you you know you are getting charged a fair bit right so make sure that you are building what you can an example is um get. com or I mean a better example is um trome.
