let me show you how you can take any API and use the new AI assistant within our connector Builder to create a custom API connector in literally minutes for this demo I'm going to be using the Polygon API it's got a pretty good API and with

our new AI assistant it's going to be pretty easy to plug in and create a connector with this so let's get going I've already created an account and I've got my API key so what I'm going to go ahead and do is go over to their

documentation and now that I'm on the documentation what I'm simply going to do is I'm just going to copy the URL here and switch back over to Airbyte now that I'm back into Airbyte let me go over to the left hand side on the

Builder tab going to click new custom connector start from scratch and then I'm going to name this I'm going to call this simply Polygon I'm going to paste the URL that I got and now down here on the First Data stream it's asking me

what endpoint I want to use first and so I'm simply going to use the ticker data stream now before I click create what the AI assistant is going to do is it's simply going to scrape the URL and plug in all the necessary information we need

on the connection Builder side to run and use a custom connector so let me go ahead and click create and now AI is going to do its thing it's going to scrape the documentation grab all the necessary information like I said and once we get back into the connector Builder we'll

see that we'll have the information that we need filled out so once the AI assistant is done scraping through the documentation you're brought back into the connector Builder what you'll see here at the top is it noticed what the base URL is as well as the type of

authentication needed in order to interact with the polygon API in this case we need a bearer token and on the top right here you can see that we can input a testing value which is going to be our API key and when you publish this connector it's going to be whoever

the users is they can put in their own API key in order to use this connector so let me go ahead and plug my API key in and click save now if you look to the bottom left here on the streams section there's the tickers endpoint that we

used at the beginning of this video and so if I click that the AI assistant has actually detected the correct URL path what kind of HTTP method it's listed a primary key as well as even gone into the records themselves and selected the correct field path we need to tap into

to get our results which is the results field path scroll down a little bit more and it added some pagination for us so it even detected the pagination we all know that pagination and APIs are pretty difficult to work with especially since they're always different so it actually

went through the documentation found that cursor pagination is the type of pagination that it uses put in the cursor values as well as the stop condition for us so we don't have to do any work right here in the Quest options I know that it didn't list anything here

but I know that there are query parameters available. Keep in mind the AI assistant is in beta, so it may not detect everything. Just keep in mind. Now for the query parameters, I'm simply just going to add a limit parameter, and I'm going to put a 100 because I don't want

every single ticker symbol to come back from this API. I just want to limit it to 100. Now that that's in place and we added our API key, let me go ahead and test this endpoint, and boom, you can see that I get back our records here. Let me

scroll down, down. You can see that we have every single value. Pagination comes through. We can click through on the request to see what the request itself looks like and what the actual JSON response is. We go to the left-hand side again on the streams. If I click the plus

sign and click this drop down, you'll notice that we have a couple other endpoints here, and this is because the AI assistant has actually detected all these other endpoints and can create new streams for us simply just by the information it got back. So what I'm

going to use is I'm going to use the news endpoint. I just want to add another one. Let me go ahead and click create. AI is going to do its thing, and now you'll notice the URL path has changed. We have a new primary key, the

field path remains the same since the endpoint always pretty much returns back in results array with the AI assistant. I don't have to read more of the documentation. It's already done the work for me, so it saves a huge amount of time. So if I go ahead and test this as well

and now in the records I get 10 records back within a page I can scroll through all the different pages with all the different news for different companies from the API so now that I'm in a good spot what I can go ahead and do is click

the publish button at the top right I'm going to do ignore warning since there are a couple things that I haven't done let me go ahead and confirm this and I can name my connector I can give it a description publish this to the workspace so that all my workspace

members can utilize this connector if this is something that I think I would want to contribute to the Airbyte catalog as a whole I can simply do that you'd have to test all your streams accordingly and make sure there are no warnings I'm going to go ahead and

publish this to my workspace just like that that if I exit the Builder go to sources click new source and go to custom we're going to see my Polygon connector is here plug in an API key and we're off to the races if you want to

try out the AI assistant for yourself you can sign up for Airbyte Cloud which the AI assistant is available for only for Cloud you can get a 14-day free trial right now so click the link below get using it let us know what you think

Can AI Really Create Data Pipelines in Minutes!?

Can AI Really Create Data Pipelines in Minutes!?