How to Add Custom GPTs to Instagram DMs (OpenAI GPTs Tutorial)

in this video I'll be showing you step

by- step the fastest way to connect your

custom gpts to Instagram to completely

automate your DMs I'm making this video

today because Instagram can be an

incredibly powerful sales Channel when

used right and by connecting your gpts

to Instagram you can completely automate

key tasks like customer support and lead

generation for your business or for your

clients the issue is that connecting AI

assistance to Instagram DMS can be an

extremely painful process and something

that I've had constant issues with over

the past couple of months so I wanted to

create a video to walk you through the

easiest method that I've found so that

you can get things set up and avoid the

mistakes that I've made at this point in

the video I would normally give you a

big explanation of what we're building

but I think it's easier if I just show

you the finished product so we're going

to get a live demo here of exactly what

we're building in this case I'm using

the testing keyword it's going to

trigger the automation that I've set up

within mini chat which you'll see in a

second so in this case I've got some

stories up of me and Dubai if you're not

already following my Instagram it's

going to be available somewhere here you

get to see bit the behind the scenes of

my life and some of the stuff I get to

off camera um but in this case I've got

some stories up and I can reply to it of

testing and it's going to trigger the

Automation and you should see a pop up

at the top there we go so it says Hey

Liam what would you like to know about

the accelerator in this case it's

pulling the first name from my account

so if your name was Sandy or Smith that

would be using that name as the first

name um but because this account name is

Liam it's going to use Liam what would

you like to know about the accelerator

would be happy to help um who is the

owner so we can start asking some

questions in this case it's going to be

taking the question that I send there

it's going to be sending it to mini chat

and then mini chat is going to be

talking to our repet back end which

you're going to see all the code in a

second and I've templated it all so you

can just copy it and follow the steps

I'm about to show you and you'll be able

to get it up and writing so in this case

let come back Thea accelerator program

is Liam mle it's designed to help

individuals establish a profitable and

scalable AI business 90 days is there

any specific information you like to

know about the program um how how much

do the plans

cost

[Music]

the AAA accelerator offers two

membership tiers standard plan which is

$97 monthly Premium plan which is 197

which is currently sold out and so I can

keep asking questions about the program

I can ask it what are all the benefits

what do I get out of it

Etc what are the benefits on the

standard

plan

for

97

what do I

get you get access to over 15 live

workshops a month you get 24 assistant

from aam you get exclusive tools

accountability groups vets Community

templates all the good stuff so what we

can say now is to try initiate the lead

capture if I say I'd like to

join great to here I'll need your full

name and phone number including your

country code to get started so so this

is going for the lead capture now I can

say my name is um

Bruce

Wayne um plus 971 5 2 7 7

8 there we go I can send that

information it's going to pull out the

name and it's going to pull out the

phone number and it's going to send that

off to air table to log that lead so

that our sales team can follow up with

them so that's the that's the build

that's what we're going to be creating

here you can easily modify this by

changing out the knowledge base as

you'll see in the steps that are coming

in a second and you can also change the

CRM from Air table to whatever you want

but you'll be able to take the template

modified as you need and I'll be walking

you through step by step the most

valuable part of this is going to be the

mini chat template that I'm going to

give you which took me like an entire

day to create because there's so much so

many issues that I have with it due to

timeouts and stuff like this which we go

into in a sec we have a ton of stuff to

get through so let's get into it to

achieve the functionality that we just

saw we're going to need of course the

assistance API from open AI we're going

to need a repet to host that code then

we're going to call our rep code from

mini chat and I will mention it again if

you haven't seen any of my other videos

but we have the two sort of sides of

this gpts release which is firstly the

gpts on the chat GPT site where you can

create them and sort of go through them

and use them use them on the browser and

things like this and then we have the

assistance API which is really the the

meat and potatoes and how this thing is

going to be used and rolled out to

businesses and to real world

applications so while gpts are fun and

all and they're cool to create around

and sort of share around with people

most of the value is going to be created

with this assistant API so you can

deploy your gpts and your your setups

essentially to different channels in

this case we're going for an Instagram

deployment and using the GPT technology

for that on my prior videos we've done

website deployment so if you want to

check out how to put these on a website

rather than on Instagram you can do the

exact same process but put it onto a

website watching the video that's up

here I've got a couple of them now but

this is the best one to start with okay

so to get started with this build you

first need to get access to my repet as

always this is going to be available in

the first link in the description you

can go to my resource Hub sign up for

free and it's going to be at the bottom

of the list in the left panel you can

click on it there'll be a link that will

take you to this ret and you should be

greeted with a screen like this first

step is to go to the fork button on the

right we can click Fork here and we can

call it I'm just going to put a fork at

the end and then I can Fork the tripple

give it a second to boot

up then you should be greeted with

something like this for those of you who

don't know how to code don't freak out

you don't need to change any of this

really and there's only a couple parts

that I'll walk you through the process

of and that's mainly around API key so I

build this in a way that it's all ready

to go out of the box I'll just take you

through a bit of a breakdown of how this

is different from some of my other

tutorials and the way that I've had to

modify this so that it works with the

Instagram deployment this actually took

me quite a while to get working because

there's limitations of Min chat which

I'll go into later um but we can start

off where they breakdown of the main. pi

file which is where all the all the

action is happening when we're actually

interacting with this application

similar to the other builds that I've

done firstly we have the SL start

endpoint which is going to start the

conversation whenever we start a

conversation within mini chat a new

person interact with my story they hit

that trigger it's going to call this mpo

the/ start and it's going to create a

new thread we're going to give that

thread back to our mini application so

here we are giving the thread ID back

and then when the user actually asks

their question it's going to be sent to

this chat inpo and that chat inpo is

going to grab the thread ID and the

message then let's go to start a run

what's called a run in the new open AI

documentation and the way that they've

set up the assistant API so you have

what's called a run and in this

particular build we have split the chat

endpoint between this chat one and then

the second one called check now the

reason we've done that is because of the

Timeout on Mini chat is 10 seconds and

when I was testing this before I wasn't

able to get responses on Mini chat

because it was running over that

10-second timeout when I was making the

API Co so we've had to split it into

these two different parts this chat

endpoint is just going to create the run

and sort of send our information and

what we need to do up to the open AI

servers and it's going to be like okay

that's what we want now give us a couple

seconds and we'll wait for it to come

back and the way we actually grab the

response from that endpoint and we've

called the open AI API to give us our

response then we go down to this check

in point we're going to be periodically

checking in to see if our process our

our operation has been completed and if

it has been completed then we can pluck

it out and send it back to the user so

might sound a bit complicated but in

this case it's going to make a lot more

sense when we get into the mini chat

side of thing so bear with me next I'll

take you to the functions. pi file this

is where a lot of the the core

functionality is happening um in this

case we have the lead capture so this is

a air table lead capture we going to be

replacing this URL with a new one um as

you set it up on your end we have API

Keys which we'll be setting up in a

second this is the logic for creating a

new lead in the air table as I mentioned

and this is the create assistant

function which is really the main bit

that you need to wrap your head around

before going out and trying this on your

own so the create assistant function is

because when you're trying to create a

new GPT programmatically in this case

we're not using the GPT Builder we are

creating our own programmatically we

need to give the instructions or the

ingredients to the open AI API to create

our new assistant so in this case we

have of course the knowledge component

which is going to be a document here on

the left called knowledge. dox then we

have the instructions which are going to

be in the prompts dop file which we show

you in a second then we have the

functions which is set up here so I'll

show you how to change each of these

around so that you can modify this so

that you can create your own gpts for

your own use case so in this case we

have the file here which is going to be

the knowledge based document if you want

to change the knowledge base to your own

document you need to delete the

knowledge based document on the left

here and then click upload file here

upload a new file and then you need to

change the name of this file here from

knowledge. do to whatever it is in the

left panel so name here should match the

name of the document on the left hand

panel next if you want to change the

promp thing or the instruction for your

gbt it's going to be stored in this

variable here so the assistent

instructions is imported from our

prompt. piy file which is over on the

left here so if you want to change the

promp thing or the instructions for your

GPT click on the prompt stop High file

on the left I've taken this out because

it doesn't look very clean to have this

massive amount of text sort of stuffed

into your functions of P file so I've

moved it out to here so you can go in

and modify this as you like and any

changes that you make here are going to

be automatically pulled in to this

create assistant function when you

create it to to reiterate we have the

knowledge component here which we can

change we have the prompting component

here which we can change via the prompts

dop file and then finally we have the

tools section uh now the tools and order

to be able to use I should have

mentioned this in some of my other

videos so thank you for those of you who

pointed it out but the model here that

we're calling the GPT 4 preview you

actually need to have a card put on your

open AI account and you need to add some

credit to your account if you haven't

yet done that going to need that which

will show you in a sec but in order to

be able to access this new model which

is able to use GPT the much larger GPT 4

which is faster and and better in a

number of ways we need to do that step

so just to be clear some of you had

issues with that last time that is the

reason why and I'm going to be showing

you in a second how to fix that so when

it comes to tools this is the more

difficult part to change around I will

be completely honest but just to walk

you through it firstly we have the

retrieval tool here this is the

knowledge base as you can see I put a

lot of comments throughout all this

document so if you get stuck you have a

question about something not working or

you want to know how to modify things

take a look at these green comments and

you'll get a much better idea of how

this application is set up so we have

the retrieval tool which is the

knowledge base this needs to be on for

it to be able to access that knowledge

document then secondly we have just one

function here which is the lead capture

the lead capture in this case is using

the create lead function which is up

here it's referencing that name then we

get to the two parameters required to

call this function firstly we have the

name which is obviously the full name of

the lead and then we have the phone

which is the phone number of the lead

including the country code it's

important to know that to operate this

application when you have the three

different ingredients you have your

knowledge base your instructions then

these and then the tooling here as well

when you go to create an assistant when

we run this app up here for the first

time it's going to look on this left

panel and see if there's an assistant.

Json file in this case you can see that

we don't have one so it's going to

follow this and say okay if there is an

assistant. Json file let's load that but

if there is not then let's go through

the creation process and then create a

new file down here so this is our way of

of not duplicating and making a ton of

different assistants when we don't need

it you're going to be charged API usage

by how many assistants that you have up

especially if you're putting a lot of

knowledge in them you get charged by the

the amount of storage that you're taking

out with your documents as well so in

this case it's only going to create a

new assistant when there isn't one

already in the side file and this also

means if you've already run the app once

and you have a assistant. Json file on

the side panel here if you want to

modify the knowledge the prompting or

the tools then you need to delete that

file and then run the app again it will

recreate your assistant. Json file with

a new set of settings okay and with that

orientation out of the way we can get

into modifying the very few parts of the

code that we actually need to change in

order for it to work on your system

firstly we have our open AI API key

pretty straightforward one you can go to

platform. open

a.com then on the left panel here you'll

have an API key section and you can

click here and create a new secret key

it's going to pop up with a a key that

you can copy and you can copy that you

can take it back over to reflet here

then you need to go to the secret

section down the bottom left we can

delete this Google Cloud API key cuz

that's not needed for this one um but in

this case we have open AI API key you

can edit this and then you can paste

yours in and then update the secret and

as I mentioned before in order to make

sure that you get access to these newer