exactly 2 months ago I had made a video on how to build an AI WhatsApp chatbot without needing any sort of code which helped many of you guys out but today I will be showcasing a more streamlined and faster approach in creating various types of AI chat Bots within WhatsApp you may be wondering what's the point of implementing AI into WhatsApp particularly well there's actually a lot to it many businesses operate their customer service solely through WhatsApp not just customer service but many different business operations and then you have a lot of regular users who
just utilize WhatsApp for regular chatting and this is where AI can be implemented to enhance many of the chat components but basically for the business aspect by creating a 24/7 support service with the help of AI we can provide fully autonomous support to various functionalities one main reason why an AI chap out within WhatsApp can streamline business operations is that you can utilize it for booking appointments order tracking email transactions as well as troubleshooting which would free up human employees from mundane task in essence it's basically implementing AI to enhance and automate many areas which
is going to reduce costs it's going to increase efficiency and it's going to improve productivity to help us accomplish this goal we're going to be utilizing an AI automation an AI chatbot creation tool called Vector shift Vector shift is an an amazing easy to use AI automation platform that requires no code at all it has a drag and drop UI to help you basically build out AI Solutions such as AI search engines AI assistants AI agents chat Bots as well as various other AI automations I've made multiple videos on this in the past where I
automated my emails slack channels and various business operations definitely check out those links in the description below cuz there's full-on tutorials which showcase how you can aut autate almost anything but with that thought guys let's get straight into the video and showcase how you can Implement AI into WhatsApp so now let's get started what I want you to do is head over to Vector shift's website which I'll leave a link to it in the description below and I want you to click on the get started button over here once you click on this it'll prompt
you to create your account either with your email address with a Google account or with GitHub so simply just sign up and we'll proceed forward once you sign up or log in you'll then be sent over to the main dashboard of vector shift this is where you're going to be able to manage all your chat Bots and various AI automations that you can create with Vector shift you have a Marketplace in which you can access readymade templates to various sorts of Creations that could automate basically like a Blog Post article generator you have different chatbot
templates that you can adopt and you can see that there's a wide variety of different templates that you can get started with right away you can upload your own knowledge base various file types manage your automations chat Bots forums or even voice spots and the best part is you're going to be able to track all of your different automations or chat Bots with the analytics now what I want you guys to do is click on this addition sign which will prompt open this window where it'll tell you to create various sorts of things you can
create a pipeline from scratch you can create a knowledge base chat Bots automations as well as various evaluations to track your automations in this case we're going to be creating a fullon pipeline from scratch and like we saw in the marketplace you're going to be able to access these readymade templates for various sorts of Integrations and various sorts of categories like assistance for productivity for finance and strategy for your Gmail and so many other options but let's go ahead and create our pipeline from scratch once you click on this button it will then take you
to Vector shift's no code Builder which is a drag and drop Builder that will help you in creating your automation or your AI WhatsApp chatbot now last time I did a video on an AI WhatsApp chatbot I showcase the basics of setting up a chatbot but this time around we're going to be going further in depth and making a more easier and streamlined approach to creating this chatot it's going to be a bit more intricate but it's going to be easier than before now what I want you guys to do is start off by placing
an input node and an output node for queries to come in and have them output it now we're going to be creating this AI WhatsApp chat up for an e-commerce site where I will be showcasing how you can integrate a knowledge base to answer customer queries next I want you to go over to the large language model setting and place in an open AI large language model node you can choose various providers in my opinion open Ai and anthropic have the best LMS out there so you would want to choose between these two in my
opinion I'm going to be choosing the gp4 Omni and I'm going to be giving it a system prompt now like I mentioned this is going to be a AI chat bot for uh e-commerce store and it's AOS e-commerce brand and the system prpt that I'm giving the open AI large s model node is to classify whether questions is about a product or pricing when they are being inputed into the chatbot and basically we're going to have it so that there's going to be a logic node a condition node which will process these two different categories
pricing and product and this way it will work its way to answering those questions based off the condition now the reason why we have this structure is because imagine we have this data set like a CSV that has a table that has listed all the different types of products as well as the prices for these different products the AI actually does a better job in retrieving these answers based off the queries from this table than a knowledge based would which is why we're going to be going through this condition approach and this is why we
wanted to root it differently this time around based on the types of questions now to process the question for these queries we're going to be placing in a text node and you can do this within the general tab we're going to Simply rename this as the question so from here we're going to then give it a text to describe what this question node is or this text node so I gave it this prompt or I would say this text you receive a product catalog with columns the product ID which we see in our data set
the product name the product brand the gender price and Etc to find the product or similar products use columns product name and description so I'm giving it an instruction or a prompt to basically reference these two categories where it's going to be processing through the question now just note guys it doesn't have to be this intricate you can obviously make it more basic where you just have a simple input node connected to an opening at large Nish model node which is going to be processing all the queries and you can have a knowledge base which
you can go paste in your different files so that it answers questions based off the context that you give it this is a simple flow where you can connect these four nodes together and have it connected with the next step this rough example that I'm showcasing right now is just how you can simply set up all these different nodes to create an AI chatbot that can process queries for an e-commerce store so essentially what will happen is a query will come in about a particular type of question regarding our e-commerce store whether that's related to
the product or the price and from here this condition node will basically set ction of this condition if it is a product it'll go through one of these other nodes and if it is about a general question it'll reference the knowledge base so we can add this condition and we can have it so that if it is not about these two different topics we can then have it go through the next step which is where we told it that I'm unable to answer this question so if it's a general purpose answer or question we'll have
it reference the knowledge base cuz this is where it does great in terms of answering questions with text whereas if it is about a price or a product it'll reference the next stream which is going to be where it will reference our CSV file so now what we're going to be doing is having the input connected to the question and we're going to have it so that this knowledge base for the general query is going to be connected over here now you can create your own knowledge base give it a name in this case we'll
name it Fitness which is going to be the type of products that we're selling you can uh enable Advanced document retrieval as well as hybrid search and you can simply click on Create and you can have it so that you can upload your various file types you can also have different Integrations integrated within your knowledge base with various sorts of thirdparty applications and this is the flexibility that you get with Vector shift also the ability to script URLs and creating an uploading folders next we're going to go place in another large language model node and
we're going to be placing in another system prompt for this and the reason why is because it's going to be answering questions for the knowledge base and this node will process all the general types of questions that are asked for our e-commerce store you can see that the job that I gave it is to answer questions using context and the conversation history next I want you to go over to the chat Tab and place in a chat memory and you want to basically now configure your prompts to connect all these notes the question being inputed
it can be linked back to this open AI large neish model node which is going to help answer questions by the way to create these different variables you want to click on this button which is to insert variables give it a name and this will basically help you connect all these various nodes so now the chat memory is going to be linked to the history and the context will be linked to the knowledge base so now that we have finished the first part of our condition which is to answer general questions referencing the knowledge base
we can then focus on the next bit which will answer questions based off the CSV so what I want you to do is connect these simple nodes which is to connect the condition to the response what I want to do is also have queries INF flowed into the prompt which is the open ALR sish model node and we also wanted so that this input which is going to process the condition of products within the CSV will get processed through this text node next I want you to place in a CSV query loader this is going
to to be loading data from our source which is our CSV file and I also want you to place in a file node where you can upload your files so in this case we can upload this data and it will process through the conditions that we have set and it's going to then go through and answer questions based off of what we had specified now I also went along and I've connected everything and spaced everything out so it looks a little bit more appealing lastly what you want to do is paste in this merge node
you can go over to the logic Tab and place in this merge node which will basically merge the two flows together so you can get the response from the open AI large neish monol Lo connected to this node and you can also have this node which comes out of the condition which is the last one that we have from this condition based off the requirements and you can have that connected and then you can have the CSV uh query loader connected to the merger and from here you can then connect the final bit to the
output node and now you're basically ready to deploy changes so to test if this is functional I asked what types of shirts are for men and you can see that it's going to reference this product log or our CSV file and it's going to be giving you answers based off of that log that we have you can see that it has a description showcases the name as well as the brand in certain cases the next step is to now deploy this within WhatsApp so now that we have our AI chatbot created you can then click
on export pipeline as a chatbot and you can give this a name anything you would want let's just name it Ecom let's click on create chatbot now I know this is not for WhatsApp but you can basically style this so that you can then export this as a link and you can even have it so that it could be protected with a password you can integrate it within your website but now what we're going to be doing is going to integrate this within WhatsApp the next thing is is working with twillow so now you're going
to need to basically create an account with twillow it's completely for free this is a Communications API platform that you can access right now and I'll leave a link to this in the description below so that we can basically connect it with WhatsApp so once you have created an account we can then proceed forward so after you create an account it will then sent over to this dashboard of toow you're going to be able to see your account information which is going to show the Sid and the authorization token which going to need so copy
the S ID go back into Vector shift and paste it in for your a token go back into toow copy it and then do the exact same and paste it in now go back into toow and I want you guys to click on messaging now what you can do is you can try it out and click on send a WhatsApp message once you do so we're going to be then acknowledging and agreeing based off of the uh terms and condition do it at your own discretion and then you want to click confirm next up is
having what WhatsApp web or the application itself on your computer opened up so you want to make sure that you have it opened up by simply testing it out and then you want to head over to sandbox setting now what you're going to do is get back into Vector shift itself the main dashboard and what you're going to be doing is connecting the API to this URL that we see over here we're just going to be placing it in over here so to do so you want to head back into Vector ship's dashboard you want
to click on the drop down menu and you want to click on usage this is where you can head over to the API keys and you can basically generate your own API keys so after you copy your API key you want to head back over to this tab within Vector shift and you want to paste it within this area so you want to make sure that you remove the brackets as well and then you want to paste in that API key and then after you want to copy this overall key and you want to go
back into twellow and paste it within this section over here and then you want to click save and that's basically it now you have an AI WhatsApp chatbot and you're going to basically have it so that it can provide relevant answers based off the context that I gave it this is just a rough example of what I was able to generate I asked it what are some good men shirts and it gave me the brand the product name the price as well as the description and you can see it also provided me a lot more
in terms of the different options it has I definitely recommend that you check out the vector shift documentation cuz it is going to Showcase how you can export your chatbot further in detail as well as customizing it meaning that you can fill out this form over here to have it so that you can connect your chat bot to an official business profile meaning that it will have a business Outlook it won't be the basic pillow Outlook with that logo it would be based off your own logo as well as your own preference and styling and
that's basically it for today's video guys I hope you enjoyed it and you got some sort of value this is how you can easily integrate AI within WhatsApp and this is just the base line of it there's so much more to it you can add automations within what uh WhatsApp you can add appointment senders you can have automations that can fulfill tasks within WhatsApp with the help of vector shift so I definitely recommend that you explore this even further so that you get a better idea but that's basically it for today's video I definitely recommend
that you try Vector shift cuz this is an amazing platform that has automated so many things for me but that's basically it for today's video guys make sure you take a look at the patreon so that you can access our private Discord as well as different AI subscriptions for free make sure you follow me on Twitter a great way for you to stay up to date with the latest AI news and lastly make sure you guys subscribe turn on notification Bell and like this video as well as our previous videos so you can stay up
to date with the latest AI news but with that thought guys have an amazing day positivity and I'll see you guys fairly shortly peace out f
