The Chatbot Bootcamp : Build Your First Chatbot! | Sachin Bhatnagar | Skillshare

Playback Speed


  • 0.5x
  • 1x (Normal)
  • 1.25x
  • 1.5x
  • 2x

The Chatbot Bootcamp : Build Your First Chatbot!

teacher avatar Sachin Bhatnagar, Dad and JavaScript guy

Watch this class and thousands more

Get unlimited access to every class
Taught by industry leaders & working professionals
Topics include illustration, design, photography, and more

Watch this class and thousands more

Get unlimited access to every class
Taught by industry leaders & working professionals
Topics include illustration, design, photography, and more

Lessons in This Class

17 Lessons (1h 26m)
    • 1. Introduction to ChatBots

      8:42
    • 2. How do Bots Work

      8:35
    • 3. Introduction to BumbleBot

      2:27
    • 4. Setting up a Facebook Page

      3:28
    • 5. Chatfuel - A quick tour

      9:15
    • 6. BumbleBot Says Hi - Introduction

      2:46
    • 7. Downloading Assets

      0:32
    • 8. Setup and the Welcome Message

      8:29
    • 9. Favourite TV Show

      8:08
    • 10. Can you guess my favourite show?

      4:40
    • 11. My favourite movie

      4:41
    • 12. Ask a question

      7:17
    • 13. BumbleBot Says Hi - Marvel Facts

      5:19
    • 14. HelloBot - DC Facts

      3:43
    • 15. Handling Small Talk

      4:11
    • 16. Summary

      2:42
    • 17. Congratulations

      0:36
  • --
  • Beginner level
  • Intermediate level
  • Advanced level
  • All levels
  • Beg/Int level
  • Int/Adv level

Community Generated

The level is determined by a majority opinion of students who have reviewed this class. The teacher's recommendation is shown until at least 5 student responses are collected.

146

Students

--

Projects

About This Class

f511c3a5

Chatbots are computer programs that can interact with humans through a simple conversational interface. They are designed to simulate an interaction with another human. Couple this with the fact that more than 90% of smartphone users spend most of their time in messaging apps such as Facebook Messenger and you have an excellent opportunity to build highly interactive chatbot based services that can empower business and enterprise like never before.

Chatbots have emerged as fantastic tools for marketing and productivity which is why companies from Amtrak to Fandango and many more are embracing Chatbots and getting a massive boost in their business through pure customer delight, that translates into conversions.

In this first class of a multi-series course, we will understand what a chatbot is and then we'll build our very first and basic chatbot for Facebook Messenger. Throughout this course and subsequent classes in this series, we will be using Chatfuel, one of the most popular bot building platform for Facebook Messenger. 

In subsequent classes, we will gradually up the ante to create advanced chatbots that incorporate features such as Natural Language Processing, Databases, Webviews, Geolocation and more.

Prerequisites

  • No coding skills are needed for this class.
  • You should be familiar with using Facebook Messenger.

Who is this class for?

  • Anyone who wishes to learn about conversational UIs and chatbots.
  • Marketing managers, Advertising and PR execs, seeking to¬†create chatbots for their clients.
  • Web developers, Mobile app developers who wish to learn the basics of building chatbots.

Meet Your Teacher

Teacher Profile Image

Sachin Bhatnagar

Dad and JavaScript guy

Teacher

Hello, I'm Sachin. I began dabbling with Computer Programming & Graphics at the age of 14 on a Sinclair ZX Spectrum computer that ran the BASIC programming language.

During the early 2000s, I forayed into Computer Graphics & Visual Effects training and production. From developing world-class curricula to imparting training to over five thousand students in the classroom and over twelve thousand online, I've actively contributed to fuelling innovation, creating brand identities and crafting innovative software solutions for corporate companies and individuals alike.

 I've also supervised award-winning animation and visual effects short films and has worked with one of India's topmost Media & Entertainment Education brands as a Vice President.See full profile

Class Ratings

Expectations Met?
  • Exceeded!
    0%
  • Yes
    0%
  • Somewhat
    0%
  • Not really
    0%
Reviews Archive

In October 2018, we updated our review system to improve the way we collect feedback. Below are the reviews written before that update.

Why Join Skillshare?

Take award-winning Skillshare Original Classes

Each class has short lessons, hands-on projects

Your membership supports Skillshare teachers

Learn From Anywhere

Take classes on the go with the Skillshare app. Stream or download to watch on the plane, the subway, or wherever you learn best.

Transcripts

1. Introduction to ChatBots: Hello there. Welcome to the chat board boot camp, where we will learn to build chat boards for Facebook Messenger using chat Feel, an online service that makes it really easy for anyone to create a chat board for their business. This is also the very first class in the chat board. Would camp Siri's very well learn toe build your very first Chad bark? But before we get down to the business of creating actual chat boards, let's get to the basics. What is a chat board? A Champ part or a chatterbox? Is essentially a computer program that can interact with the human through a conversational user interface there, also known as conversational agents or dialogue systems and are meant to simulate a human in the best possible way. Ah, conversational user interface is simply an interface where the user interacts with the software in much the same way as a regular conversation in a messaging app like Facebook Messenger or WhatsApp. If you've used Siri or Google now on your smartphone, you've technically interacted with a chat board, although there strictly Weiss based virtual assistance. What we're talking about Instead, our text and Vigee it based conversational interfaces like the one you see on the screen. We can check with the board in much the same way as we do with a friend or a coworker, and it responds back in a human like tone that excuse the personality that the board developer has designed. Additionally, the board also offers easy access Barton's veg, its galleries and other visual elements that make it really simple and fast for users to interact. This is such an exciting time to be in when computers are learning the art off human expression and conversation. Of course, Behind the philosophy is a maze off algorithms and technology that is bridging the gap in human computer interaction. But chan ports and conversational interfaces are not new. In fact, the very first chair port was made back in 1966. It was called Eliza, built by Joseph Poison Bomb at the binding M. I T. Artificial Intelligence lab. Eliza was made to recognize human interaction by way off simple pattern recognition. It was released to the administrative staff at the lab, in the guys off, a psychiatrist wise and Bomb was shocked to observe that people actually thought that Eliza was a real doctor, and they instilled so much confidence in the chat board that they even discussed personal problems with it. As he notes in his book Computer Power and Human Reason, Joseph Reason Bomb was both shocked and felt rewarded at the way humans behaved with a computer program. Users felt emotional when chatting with Eliza, which in itself is incredible. This was also the first chat board to win the Buring test, which was developed by Alan during in 1950 to judge a computer's ability to display intelligence to the point where it becomes indistinguishable from human. In fact, Eliza is so popular that it still lives on on its website and is quite popular even today. What elicit darkness was the fact that humans have a desire to communicate with the computer in much the same way as they do with another human. From 1966 Let's fast forward to 1995 and the boat off Alice or the artificial Linguistic Internet computer entity inspired by Eliza. Alice also employs pattern matching to process human input. Alice also introduced the artificial intelligence markup language, which has since being used on several other boards as well the year 2006. So IBM's Watson Ah computer system designed to answer questions. Actually, when the popular quiz show Jeopardy, Watson used natural language processing ah, subset off artificial intelligence that deals with interpreting and processing human speech . But Jack ports always remained within the debian, off tech labs and large organizations where they were essentially research projects ever. Never really consumer focused, By the way. The closest alternative that we've had to conversational chat parts has been the interactive voice response system, or idea that we've been using for years. Now. Most large organizations provide ah pre recorded waste based three off options when a person caused their official number. Callers then have to navigate by pressing numbers on the keypad. That sends a DTM F tone, which signals the I Veer system to perform a certain action our present. Yet under the menu off options, you find idea systems everywhere, from banking to customer support and more. Needless to say, they're not as intuitive or user friendly as one might think, and the introduction process is primitive at best. This is what chat parts after they have an opportunity to change, probably the first ever real consumer focused product in the space was Apple's City, Ah, voice based virtual assistant that employs artificial intelligence to provide intelligent responses. Introduced in 2010 Syria allows users to literally talkto the smartphone. This was followed by Google, now Microsoft's Cortana, Amazon's Alexa and Google Home Products. All these devices and software are Weist based watching assistance since the court technology that drives all these products is artificial intelligence and, more specifically, natural language processing. There has been an upsurge in the chat based conversational agents space as well, and this is the subject of this course. But the real push towards chat boards came in early 2016 when Facebook announced its messenger platform, which allows developers to hook in and create check ports. And while other messenger platforms have had their board channels as well, Facebook's product has been seen as a game changer, goingto over 900 million people who use the messenger platform. That is a huge audience to cater to, and it is not surprising to note that Facebook's move has caused other companies to take note and invest heavily in the space. Of course, not everything goes as planned in March 2016 Microsoft launched a chart board called Day, which is an acronym for Thinking About You. They were supposed to mimic the personality off a teenaged American girl, and it was supposed to interact with people. On Twitter, they incorporated artificial intelligence and machine learning. What this means is that they was designed to learn from the replies that got on Twitter and self augment its responses accordingly. Humans can be cruel. Based on the responses that they received. It started an unfortunate spew off racist and offensive tweets that forced Microsoft to take it down in less than 24 hours of existence. The silver lining here lies in the learning outcomes and a test of capabilities of chat boards. And as they came too early experiments and rocketry. Many failed but ultimately went to the moon and beyond. This is what makes chat ports super exciting there rapidly trading places with traditional more by laps. And while that doesn't hold true for everything, ah, lot off online services that required a dedicated app can now be offered as a chat board on platforms such as Facebook Messenger. This also comes in light off the fact that For the first time in the history of smartphones and APS, people are spending more time in messaging APS than even on social networking apse, which earlier dominated the industry. This is a mind blowing fact and has been a catalyst in the push towards chat boards. Today you have chat boards on Facebook, Messenger V Chat, Telegram kick, Skype and so many other platforms. These social media platforms are the best places to deploy boards for most businesses because this is where your customers are, and you want to make it easy for them to interact with your business through your chat board to build airports. You have services like chat fuel, dialogue flow with Dorte, and many more in this course will be specifically focusing our attention on Chatfield to build our boards for Facebook Messenger. What is Chatfield and how does the Chadbourne work? Let's find out 2. How do Bots Work: In the last lecture, you were introduced to the world of chat ports. Let's continue that streak forward and understand how chat boards actually work. We have established the fact that chart ports are computer programs that can interact with humans through a conversational user interface, which is more natural and is easy to comprehend. But how do these boots work? Categorically speaking, we've got two major types of boards, and we're talking about how these boards work at comprehending human import and processing it into a response. First up, we've got the rule based chat board. A rule based board is very implement. A flow chart off rules that define how are bought will respond to user input. For instance, here's a sample conversation flow to consider. Imagine for a moment that our chat board is designed to offer weather reports. We begin by presuming the fact that our users import message could be in one off the falling to former's. If the user says something like, what is the weather like in Hong Kong, then he's requesting current weather conditions in Hong Kong. If this is the case, then our Bart is instructed to find out the current weather and route the conversation accordingly. The board in this case is designed to find the words. What is the weather like? If this sequence of words is present than the Barton knows, it has to find the current weather conditions for a given location, and it receives to extracting the location, which in this case is Hong Kong. Once that is done, the board can query and online weather service to get the data and can then reply back to the user. The second kind of question that the use of might ask would be Village rain tomorrow in Hong Kong. In this case, the user wishes to know forecast conditions. The board sees this partner in the question and understands that the user wishes to know forecast conditions. Instead, we've set clear rules here that are based on the user's input message and the presence off preset patterns in the text. Now, for matters of practicality, platforms like Facebook Messenger also allow us to incorporate buttons and visual elements instead of making the user type in the entire thing. So the same conversational flow can also be represented This way, we're represent a series of buttons for the user to choose how he or she wants the conversation to continue, though it may seem less off a jack board where a stress on the world chat. This is how most boards are built these days because at the end of the day, these buttons offer a clear and prompt way for the user to converse with the body. In this example, the user is presented with two buttons, one East for the type off weather report that the user wants thereafter. It's simply a matter of getting to know the location for which the report is being sort and then providing the actual data in both these instances, a clear flow chart off rules and simple and easy options for the user defines how the conversation is supposed to float. Consequently, a pure rule based board has limited flexibility by itself. Humor recall the case off the interactive voice response system or the A V R, which I had mentioned in the last lecture. If you think about it, it functions pretty much the same way as a rule based Bart and I've ER is in fact a rule based response system where you press the buttons on the phone's keypad as a way to provide input. You program the rules. The system strictly adheres to the rules and creates the conversation flow. Chat boards, on the other hand, can make this process more meaningful, personalized and extremely intuitive. Now, in addition to the simple rule based board you have bought that utilize artificial intelligence now, artificial intelligence by itself is a very vast field. Boats that used E I rely on natural language, understanding and processing. Emmylou or NLP is a subset off artificial intelligence, and it deals with interpreting human speech in either voice or text and extracting meaningful data that can then be processed. The most common type off A baseboards are very trained. The system on a wide gamut off use of messages and questions. Training refers to the process where the AI system ingests user messages and, with our help, transforms that into actionable leader. We had the system understand a variety off user messages, and the system gradually becomes smarter and starts making deductions automatically. Imagine a new office worker who has no idea about his duties on the first day of his job, but the same guy after about a week off guidance and training becomes smart enough to make his own decisions. That is exactly how a natural language processing based system works in more ways than one this kind off abort. Also, borderlines are rule based board, but it's sophisticated and way more flexible and advanced in how it understands user input in such a boat. We first trained the system to understand a wide gamut off user messages, all of whom can be mapped toe their corresponding intent. The intent here refers to work the user wants at the end an intention. So, for instance, when a user say's What is the weather like in Hong Kong? Or when he says, How's the weather in Hong Kong, or is it sunny in Hong Kong? In all these cases, the intention off the user is the same. He wishes to know the current weather conditions. You can provide endless permutation off user import and have the system map toward given intent. This way, the system learns, becomes intelligent enough to reduce the right intent, even if the messages have variations that you have not trained. The system on this is where natural language processing excess, the second part of this equation is the data that is contained within the user's messages. So in the phrase, what is the weather like in Hong Kong, there is one more piece off information that is valuable to us. The location Hong Kong is required to fetch the actual weather data. This is known as an entity in this last washing off the user's query. We also have an expected weather condition, and this can be extracted as an entity as well. There are lots of ways by which these can be extracted, but using an enemy you based system makes it intuitive, smart and easy for us to extract is later, the system learns to identify these entities and with sufficient training, becomes very efficient at this task. Once we know the intent and the entities, we process a suitable reply and send back to the user. For instance, in the case off our weather board, I know that my user is asking for the current weather conditions in Hong Kong. All I have to do is fetch the weather conditions from the weather service such as the Yahoo Weather FBI and create a response which carries the weather report that the use of Sikhs. It is important to know that the responses are programmed into the system and can go from simple to complex, depending on the nature of the chapel within the domain off A. Our boards, which employed deep learning toe, actually create responses rather than fetching them from a pre programmed set. These kind of Bart's are not very common as yet, but are the apex off what artificial intelligence can allow us to do? And quite the future? The recent Google duplex demo at the Google I 0 2018 event saw a futuristic Google assistant understanding a seemingly impossible nature off conversation and generating an accurate response in a very human like way. The boat was ableto talkto a human with all the nuances off an actual human being. That is the future of boards in general beard, voice or text. I hope you now have a broad understanding off our boards work. Let's see what Jack Fuel is all about. 3. Introduction to BumbleBot: all right. Now that you haven't overview and a basic understanding of what chart fuel is all about, let me introduce you to Bumble Board in this course will be building several jack boards for Facebook messenger. Every Facebook messenger chair port must be attached to a Facebook page. You see, at this time, Jan boats are only available for Facebook pages, so you need to have a Facebook page in order to create a childbirth. If your business already has a page that would work perfectly fine as well. Let me explain how this works. Every Chatfield board or, for that matter, any chart bought that you build using any platform or mechanism must be connected to a Facebook page using a messenger platform app. This APP arts like an interface between your board platform such as chard, fuel and the messenger platform itself. Messages that the user sends to our page using messenger are related to the board platform , which in our case would be Chatfield Once are bought on Jack Fuel has a response. It is sent back to the Facebook app, and it goes on to show up in the users chat log on Messenger. The good thing here is that you don't have to write any cord to create this connection between Chart Fuel and Facebook Messenger, and this process is as simple as Click and connect. The bottom line here is that you need to have a Facebook page to begin with. Now I'm sure you know how to create a Facebook page, but if you have never done so, the next lecture walks you through the process, so there's nothing to worry on that front now. Remember I mentioned Bumble Board? Well, that is what we call our chat port. In the scores, we make a page for Bumble Board, and all examples would use the same page. From example to example, Bumble Board will assume different roles and conversational flows, but we'll keep the page and the name of the board same. My only reason for doing this is so you don't have to create separate pages for every example board that we create in the scores, and he can just focus on the task at hand building chair ports. So let's begin this by creating a Facebook page for Bumble board. If you already have a page for your business, you can use that instead as well 4. Setting up a Facebook Page: So as we discussed in the last lecture, ah, chat port on Facebook needs a Facebook page to connect. Oh, this is very important and is the first step towards building a chat board. Now, if your business already has our Facebook page, then you're good to go. However, if you're new to Bages, then I'll show you how to create one. As you know, by now we're calling our chat board Bumble Board. Every chat boards section in this course will use the same Facebook page and share port name. This will avoid unnecessary work and will save you time as you practice and build box now, since we're going to create a fresh page for Bumble. But I've included sample profiling cover images to get you started with quickly, you can download and extracting archive containing these images from the link given on the screen. Once you have extracted the images, head to facebook dot com. Forward slash pages forward slash create. You need to be loved into Facebook to access the speech. Once on it, we'll start by clicking on the get started button for the business or brand option. This depends purely on the nature of your business or the intention behind the page. If you're setting up a page for a cause on non profit or a celebrity, for instance, then you can pick the community or public figure option to begin with. I'll begin by naming the page Bumble board. Since you're building the page barely to host our demo, Bart's for the category. I said this to, ah, product, and I'll get the nearest option as product slash service with suits. Our case. Well, I love Click Continue, and I'm now required toe upload the profile picture. So let me pick the profile picture file and it gets uploaded in the background. Once done, it asks me to upload a cover photo. Let me pick that as well. After this, I reached my newly created page. Now, for some reason, my cover photo did not get uploaded earlier, so I just click on, add a cover and upload that again. There is perfectly set now. Next, you notice this big blue Barton hair, but say's add a button. If you're going to host a board on your page, it makes a lot of sense to change this to a send message button that invites people to send a message, which will then be handled by your chart board. So I'll click on this button now and here I can configure it to act in a way that best suits my page. So under the contact you section, I'll select the same message option and will hit next on the following step. Select Facebook Messenger and Click Finish. That's it. Your page is now ready to be hooked up to chat fuel, so let's explore chattel and I'll give you a quick tour off the you A before we build our very first chat board. 5. Chatfuel - A quick tour: okay, Before we create our first board, let me take you on a very quick tour of jet fuel and its main sections. This is just to familiarize you with the user interface before we begin building specific conversation flows. You're on Chatfield dot com on the home page. Here you can see chart fuel in action and on scrolling down, you'll come to this section where it say's no programming required. In all honesty, you can build pretty good Bart's without needing to touch cord. But for ultimate flexibility, you will need to invest some time in writing your own court as well. Don't worry, I show you how it's done now on the face off it chat. Fuel boards are mostly rule based when you define a conversational flow and its rules and it works really well. But besides rules, chat fuel also brings with it the power off natural language. Understanding through its built in a engine, allows the board to understand what the user has written, and this is then used to route the conversation in a particular direction. So you said the rules, but the rules are driven by artificial intelligence in the background. You'll see how this works in a sharp way. Jet fuel also offers integration with done off services some directly, some through intermediate services like Zambia. So you can do things like sending your latest instagram posts every evening to your audience on Messenger. Likewise, content from your WordPress Blawg YouTube channel and lots more can be brought in as pre formatted and ready for consumption content. The good thing about these integrations is that almost all of these don't require a single line off court to Britain, and you do have an interface for where you want to extend the capabilities off Chatfield through your own cord. We'll see all of this in the course, so sit back and enjoy. Now, if you're thinking Chatfield only builds boards for messenger and sounds like a small time platform, then think again. After looking at this insanely small but important list of companies who have built their boards. Who's in charge Field? I'm sure this changes your perception immediately. As it says right here, more than 46,000 Chatfield boards are out in the wild. That is an impressive figure right there. Now the good thing is that most of chattels abilities are absolutely free to use, and this works. For most cases, however, there is a paid option as well that will discuss as we go along. I'll be using a free account in this course. However, as I just mentioned, I will discuss how the pro account works and the benefits that it brings to the table. If you're an enterprise or if you have a large audience, so how do you begin? You hit the Create Facebook part button right here, and it logs you into the dashboard using your Facebook log in. It goes without saying that you need an active Facebook account To use Chatfield. There is no separate registration or sign up needed. Once I'm logged in, I can see the boards that I've created myself in the dashboard section here, what number one is connected toe a Facebook page and hence the Eiken. The 2nd 1 however, is not connected to any Facebook page. Besides your own boards, Jack Fuel offers a gold mine off templates for ton off business and use case scenarios. These are built up conversation flows that even modify and implement or used to simply learn how specific floors may be built. We will not be using these as well. Roll out our own conversation flows from scratch. However, once you have a decent grasp on check fuel, I'd recommend going through these in your spare time. For added insight. If you end up building a part that might be useful for others, you can even submitted as a template to be featured in this gallery with your own name and credits. This is indeed a fantastic feature. I'll now click on the create from Template Barton here, and I'll select the blank board template like So this adds a blank chap ought to my dashboard. Now clicking the Spartan brings up the dashboard where we'll do our handiwork in the upcoming lectures. This first section is the automate section where you build the rules and define how the conversation is going to flow. The left hand section here is very you create blocks. Ah Block defines a single conversation sequence and may consist off text messages, images, galleries, buttons and lots of other things. Besides these jacked fuel offers, plug ins or term it uses to define the integrations that we discussed a while back, this speeds up your board design work and simplifies a lot of complicated processes that would have needed ours off custom written code. Within the automate section, you essentially route to conversation from block to block based on users import. So essentially, this is where rules are defined. You see this in action in a short way. The next section, titled Set Up A, is where you play with natural language processing. The aim off the section is really simple. You trained the chat port to understand various kinds off user messages and you create a rule. For instance, a user might say Hi. Hello. Hey, how are you? What's up and so on? And in all these cases, the objective is the same. The user is greeting the chart port and you can create a rule and direct the chat port to send back a suitable reply such as Hey there, user. I am Bumble, But I'm so happy to see you today. The AI system on Jack Fuel learns by the off training to recognize variations in how a user might respond and becomes smarter at it. Job more on this in practice. In our examples now, the people section here is available toe pro users who pay upwards off 30 U. S. Dollars per month. For this and other premium features. This section essentially allows you to segregate your audience based on a number of parameters that you might collect within your child part. This can be used to create segmented groups. For instance, Ah bought for a college or university might want to segregate users based on their expressed interest in a particular course. This then allows the organization to send customized messages made specifically for that audience quite a powerful feature indeed and very useful for business and marketing. Next, the broadcast section does what it says Once you have an audience which has interacted with your bought at least once, you can send them proactive messages such as news updates. Fresh boasts from social media offers etcetera. The idea is continous engagement and to keep a dialogue alive. Imagine having to do this, man. Believe it. Every customer that has ever interacted with you on messenger quite an impossible task When you say with the Broadcasting Board, this is cakewalk and completely automated. The next section title configure deals with configuring various features associated with your board, such as persistent menu buttons, which are always available under you I and allow users to quickly access prominent features off your service. You can also set up a payment system for e commerce boards, where you would want to sell something from within the Chaput itself. You can also add admits, among other things, the gross section offers tools to expand the reach of your chat part. Facebook offers a selection of tools that allow you to grab audiences from outside off the messenger app and website from places like your own website through the customer chat plug in the Messenger Discovery section and other things. This section allows you to set up all these features in a single, unified place. Next, the analyze section shows extremely valuable analytics for your boat. As you can see, we don't have a board at this time, so there aren't any analytics to begin with. This section is off utmost importance for continous evolution off the chat board, and I really advise you to check this once you're bored, is public and has sufficient data to play with. Finally, the upgrade button allows you toe operator account toe a premium prohibition. You can click on the Learn More button to understand some of the benefits, and I've included a handy document in the course that discusses the benefits as well. There is a pricing calculator, as you can see here, that allows you to see projected monthly fee based on the user reach that you aim for your body and before the end. There is this test, this chat board button up here, which allows you to instantly play with the board in messenger, even if you haven't created in connected of Facebook page. So this was a quick tour off chat fuel. That's all we have time for in this course. Thank you so much. Just getting. Come on, let's build our first chat board. 6. BumbleBot Says Hi - Introduction: welcome to your very first chat board on chat fuel. This first exercise in building a chart board using chat fuel is aimed at familiarizing you with the process and the workflow that is needed to build and deploy a chat board. Our first chat board is a simple guy who likes watching movies and TV shows, but instead off me explaining watered looks like Let me show you the final result. Now, when you access the boat for the first time, you'll see the get started button. This button appears when you either access the board for the first time or if you delete the conversation and access the boat again. So let me click on the get started button now, and Bumble Board welcomes me first, and it knows my name. Do it, then shows an image off Bumble Bee and continues the conversation in a pretty manner, leaving me with helpful instructions on using the body. So as advised by the board, I'll type what is your favorite movie, and the chat board tells me about its preference and liking for movies by Christopher Nolan . Then it asks me to pick a character from the given options I like model, so I'll pick Iron Man, and the boat actually shares a fact about Marvel Comics. Next, I'll ask, What shows do you like? And Bumble board tells me that it likes Westworld, But then it asks me if I can guess another show that it likes. I think I know the answer to this one soil type is a Silicon Valley, and the board agrees fully. Finally, I'll say bye and the port bids me farewell. All right to make this board happen will first create a new chat port application on chart fuel and will connect it. So the Facebook page we created earlier, then our first step would be to welcome the user and introduce Bumble board with handy instructions on using the body. So what are we waiting for? Let's get going. 7. Downloading Assets: before we build the board, you'll need to download these images, which you will use in the project since are bought. Talks about movies and TV shows. Using images in between text messages makes the conversation more interesting and enjoyable . These images are animated Jeffs or gifts that I've downloaded from jiffy dot com. So download and extract these images from the link given on the slide, then hop and jump to the next lecture. 8. Setup and the Welcome Message: I hope you've downloaded the image assets already. In this lecture, you will take your very first steps at building a chat board. I'm here on the chat fuel dashboard, and besides the boats that you can create here, you can also see a ton off templates that can be used to learn domain specific work flows and to also adapt and billboards quickly. Once you're done with the scores and have a form grasp on building boards with jet fuel, you should explore these templates as well. For now, let's get on with our own board. The first thing I'll do is click on this blank board button, which creates a new and blank board and makes it available in the dashboard section like so , before going further, I'll click here and rename this to Hello board. Then simply click on this entire thing to get into the editor section, where you will actually build out the chapel. I have my Bumble Ward page open and available by the side, and before building anything, I'll click on the Connect to Facebook Button to connect this chat board with the Facebook page. This is not immediately required, but should be done so that you can utilize the capabilities off the chat board on a real Facebook page and not in a simulated environment in the board publishing section. I'll scroll down through the list off my pages and select the Bumble Board page, which we created earlier, and I'll hit the Connect to Paige Barton to get started there. That's how simple it is to connect your board to a Facebook page now messages sent by users to bumble but will be directed at the chat full board that we're building here. Now let's go in. Do the automate section where all the magic takes place. You notice we have two blocks already built and ready for us. The welcome message block is what will start the conversation when a user brings up Bumble board for chat for the very first time. Usually the user will see an interface like this with the get started button taking on the get started button straight away in walks. The welcome message block here and the conversation sequence that you see here will play for the user. You can see we have a block off text already with some text instructions. This is known as a card. More specifically, this is a text card, which you will use all the time to send simple text messages to the user. The text card is one off several visual elements that Chatfield offers here and are the building blocks off our conversation model. Imagine these as Lego blocks for building your conversation. You put them together and chat feel plays this out for every user as he or she converses with the Bach. These cards represent the visual elements that Facebook offers as a part off its messenger platform. You'll also notice this keyword in purple here but say's first name in a double pair of curly brackets. This is a system attributes that automatically brings in the user's first name from Facebook. This allows you to personalize messages automatically and create a more cohesive and effective conversation. Shatford provides direct access to a number of system attributes, such as the users first and last name messenger I D and so on. And you can also create your own attributes to store data will see these later in the course. For now, let's edit this text card so that Bumble Board can introduce itself like So then I'll add an image card below the text card, like so, and we'll click here toe upload an animator Jiff image from the assets you downloaded earlier will select and upload the 01 Bumble Bee image like so our boat is called Bumble Bark. But we're adding this to make the conversation more interesting and video because right after the shows up, you say, uh, not quite that window. The idea is to make the conversation interesting visual and Vitti. And since we're trying our best to make the conversation more interesting, how can we forget emerges? We can add emojis to our messages for this goto emoji copy dot com, which allows you to copy and paste emojis for this message. I probably picked this winking face with tongue emoji, and I'll copy like so, and we'll simply based it twice here. Next, I'll add one more text card, which provides more useful information to the user. So here include instructions. In our case, our bought only knows about TV shows and movies so the board can instruct the user toe. Ask questions like, What is your favorite show, or what is your favorite movie to begin with, you'll notice. I've used to text cards at the end to display information that could perhaps have been clubbed into a single text card. The reason we do this is because it mimics how the typically text chat with our friends. Most, if not all people type and send short messages in quick succession. They're easier to type and read just like a bulleted list of tasks. That is why it makes sense to break down large messages into multiple text cards. This is, of course, a personal choice experiment and find out what works best for you. At this point, I click on the test this chart port up here and once I logged in to Facebook like so Bumble board will start chatting with me. As you can see here, in this case, the welcome message block is pushed, unlike if the user were to approach Bumble board, in which case the get started button will appear and clicking. When it will bring up these messages, you can see the first message here is personalized with my name that is the first name system at real connection. You will have also noticed here that all messages rolled in together. Unlike a real person who takes timeto type, the system here just dumps all messages together. Thankfully, Jet Fuel offers us our typing. Billy's simulator. Back in our welcome message block. I'll hover the mouse between the image card and the following text card. When I do so, a plus sign shows up, which allows you to insert a card in the middle. Here I'll insert a typing card. This card similar it, said DeLay, which happens when a real user takes time to type in a message. The reason why we should use this is because this makes the conversation more believable and allows time for the user to read messages as they trickle in one by one. The deal is later can be adjusted based on the amount of text that follows and the typical amount of time that it would take a user to type it in. I'm going to set it to about two seconds here. I'll add another typing guard at this point here. And since the falling texters longer, I'll assume the delay. Toby. About three seconds now, when we hit the test button, the conversation gets appended to our existing messages, but notice the typing indicators that show up in the chart now. This is our typing 1,000,000,000 action. Interesting, isn't it? You're building your first chat board with chat fuel and isn't It's quite simple. We have just built our first block off messages visually, and it's fun. Take away to grasp the feeling that a computer program is chatting with you. That's a chat board for you now. While we built this sequence of cards sequentially, Weaken also dragged them up or down to reposition as needed. The U A hair is just amazing and simple to comprehend, so but that said, let's keep building this board further. 9. Favourite TV Show: In the last lecture, we equipped our chart board to welcome the user. Using the welcome message block. We also advised the user toe. Ask questions like, What is your favorite show or what is your favorite movie when the user does so, we want to route to conversation toe a block called Favorite TV show, which will then take care of the conversation. This way we can branch the communication and routed in a direction we like. If you think about it, we need some sort of a rule, which says that when the user utters something like, What is your favorite show, or what shows do you like or simply shows you love or shows you watch only, and only then should we route the conversation toe the favorite TV show block. You'll see this in action in just a moment before we go about the business. Off favorite shows, however, let me quickly discuss the default answer block, which also comes standard when we create a new board on Chatfield. You see chat fuel offers natural language understanding to understand what the user has said, and it allows you to route the conversation accordingly. This is rule based routing that we were talking about earlier, I discussed this in more detail and how it is vastly more sophisticated and powerful when compared two simple rules. The bottom line is that the board should be able to understand what the user is saying. Now, in instances where the boat fails to understand what the user is saying, we have a fallback, a catch all mechanism known as the default answer block. When the board fails to understand, it simply plays out the default answer. So at this time, if I say what is your favorite show? The board returns the default answer and the contents off the text card that comes as a default. Ideally and in a production oriented environment, The default answer block should offer handy instructions to the user who can, worse with the board a sort of a help mechanism. Now, to keep things simple in this project, I'll change my default answer to something like No idea about that, really, and we'll include an emoji as well. Now, if I were to ask again the boxes, no idea about that, really not very helpful now, is it? So let's build the favorite TV show block and the conversation within. For this, I'll create a separate group. Groups on chat fuel allow you to organize your blocks in a more manageable way. So everything to do with favorite TV shows will be clubbed under the favorite TV show group , which I'll set like So here I create the favorite D visual block First, when the conversation reaches here, are bought begins by saying so. I really loved Vest Worlds Season one, because I bought Loves Vest will really will naturally include an image, and I'll use zero to WestWorld from our assets. Like so. Then we'll include a one second typing Billy and we lad. Not so sure about Season two, though I'm not too sure about Season two myself. And so isn't Bumble board also at an emoji, and we'll also include an image from Season two like so by now, I'm sure you're finding this process really easy now. I'll add a two second typing delay and our board says, but I like another show. Can you guess which one? At this point, we're asking the user to name a sure that Bumble Board likes. There are no hints whatsoever, correct at this point, the conversation sequence for the favorite TV show block is done. And if you were paying attention, I'm sure you're wondering. How do we make the user go from the welcome message Block to the favorite TV show block? Where's the rule that we were talking about earlier? Well, toe build this rule. Let's go into the set up A. I section like so the a section is easy to understand. Let me demonstrate. First, we need to click on the Add a rule button, like so. Then on the left hand side will need to type in utterances that users are expected to say. The really cool thing here is that you don't have to be precise in assuming what the users will say. Chad Fuels, built in natural language understanding, understands the samples that you provide and learns to recognise similar phrases. So even if there is a deviation, the board will understand what the user is demanding, and we'll drop the conversation accordingly. In this case will offer a couple of examples to train the Emmylou system. So ah, user might say, What is your favorite show or just favorite show with both the British and American spellings of the world. Though this isn't really needed, as the system is smart enough to work through it. I will also add samples like What shows do you like one show that you love shows you love shows you watch, and you can keep adding to the list. Now, when the boat comes to the conclusion that the user is saying one of these or similar phrases, we can either get the board to say something in text, or we can simply route the conversation toe a block. And that is precisely what we will do here. When the rule on the leftist matched goto the block that we have selected on the right, which in this case is the favorite TV show block. I'm sure this is simple to understand. Now, if I ask what shows do you like? Our board plays out the conversation sequence we built in the favorite TV show block like So brilliant, isn't it? Try typing in a phrase that it's similar to the samples we provided. You see our board still works as expected. All right, now, remember our chat port? Asked a question. What is that one show that the board lakes. Let's answer this in the next lecture 10. Can you guess my favourite show?: In the last lecture, we created the favorite TV show Block, which handles the conversation. When the user asks something like, What is your favorite TV show at the end of that sequence? Are bought, asks. I like another TV show. Can you guess which one? And at this point, we're inviting the user to type in the name off a show. We want to send the user toe another block, which will call correct answer, which will handle this part off the conversation. And just like the favorite TV show, Block and its rules will also build rules that directs the conversation toe. The correct answer Block in this case are bought. Loves Silicon Valley. So are a rules. Will look for phrases such as Is it Silicon Valley? I think it is Silicon Valley and more as before. I'll begin by creating a new block in the favorite TV show group, and we'll name it Correct Answer. The reason we're putting this under the favorite TV show group is because categorically were in the realm off TV shows, and it is better to organize content this way for the content of this group. Let's begin by adding an image card first and a select the 04 Silicon Valley image from the Assets folder like Soap Harry. See Jared in a rather exciting mood. I'll then add a two second delay, and we'll add a text card where I'll begin by using the first name off the user. This is done using the first name system attribute, which you saw earlier in the welcome message. Block. Using system attributes such as the first name personalizes the messages and creates a stronger connect with the user. This is ever so important, considering the fact that a computer program is chatting with the user and we must try all means to make the conversation appear friendly and personalized. So this message reads, Do you know me well? And let's also add the smiling face with sunglasses, emoji here, then a three second Billy and another text card that says Silicon Valley is my favorite show. Although I'm missing Ehrlich, Alec Bachmann was a hilarious character on the show, played by TJ Miller. P J left the show at the end of Season four, and he's sorely missed. Okay, Next, we'll add the A rules to bring up this block so click on the Add a rule button again, and we'll add phrases like Silicon Valley. Is it Silicon Valley? I think in the Silicon Valley and so on, and this should bring up the correct answer block. Let's test this out. I'll say a word that we Silicon Valley and it books notice here that the phrase would that be Silicon Valley doesn't appear in our samples for the A rules and yet chat fuels. Emmylou was smart enough to understand that were indeed talking about Silicon Valley. Cool. Let's see what we have next. 11. My favourite movie: Now that we have handled the TV shows and the conversation surrounding it, let's get to bumble boards liking for movies. So Bumble Board likes to talk about the movies he likes. So by now I'm sure you're able to predict what's needed to get this conversation going. We lead a block, which handles this part of the conversation we call it Favorite movie. The rules for this block will include phrases such as What is your favorite movie or your favorite flake or favorite film Back on chat Fuel. I'll begin by creating a new group. This group will include blocks that fall under the topic off movies. Then, like always, I'll create a new block called Favorite Movie, and I'll begin the conversation with the text card, which say's you know, such an I. Like almost all movies by Christopher Nolan, my name would be inserted here, thanks to our good old system attributes first name. Then I'll add a two second typing Billy before the next NextCard, where we'll say I literally cried while watching into Stella. Don't look at me. I didn't cry Well, watching into Still a bumble boarded and after this, it's only app that we have of keeping Matthew McConaughey a gracing the screen. Well, not end here. We'll add another one second, Billy, and we'll say, Also, I loved inception. This will be followed by the 06 inception Image like So cool. So now that this is done, let's set up Are a rules for this block, as mentioned earlier, will include phrases like What is your favorite film favorite movie? Favorite Flake, etcetera. And we pointed toe the favorite movie block. Let's just this I'll type What is your favorite movie? And Bumble board plays out our conversation sequence as expected. By the way, did you notice something that I did differently? This time, instead of flicking the test, this chat board again Barton. I simply went ahead and typed out the freshly added phrase. And it worked. Since the board application is running on chat fuel, you don't need to refresh and reload the boat every time you make a change. That is, unless you want a sequence to be played out from the beginning. Make a change on chat fuel and directly played out in messenger. Now just for kicks, I'll type your favorite show and our TV show sequence plays out as expected. And so the question asked, I'll type game off. Thrones and Bumble Board says, No idea about that. Really. This is the default message. Block playing for you. And if I say so, What about Silicon Valley, then are correct. Answer. Block plays out as well. You can now see how the BART moves back and forth between blocks, depending on user input. Great. Now for something interesting. 12. Ask a question: in the previous lecture, we built the favorite movie block. What I want to do is at the end off the conversation in the favorite movie block. I want to ask the user a question. Since we're talking about movies, we'll ask the user to pick one of these images. The Batman image here indicates an affinity for DC comics. An island man represents a model fan. Now we can simply add this question at the end off the favorite movie block itself. But in such a situation, it is strongly recommended that functionality be broken down into manageable blocks rather than building one long series off cards this way, not only is it easier to design but is more manageable when content needs to be modified or removed. This is akin to model arising court in almost all computer languages. Ah, smaller unit off functionality is easier toe build, test and maintain than one monolithic machine. So what we'll do is build a new block for asking the question, and then, depending on the image that the person chooses, will segregate the user toe. Either a Marvel fan or a D. C. Fam thereafter will show some facts regarding the comics. But don't worry about that yet. Another question that you might be asking is, How would we go from the favorite movie block? Do the ask question block. Don't worry about it just yet. As just described, I'll begin by creating a new block called Ask Question. I lose a text card and will say such a Let me ask you a question out of the following two characters. Which one would you pick? Notice the use off the first name system attributes, which by now should be very familiar to you. Now here we want to offer two options to the user. This can be done using the quick reply buttons like So you could create two buttons 14 Option one and the other for Option two. And these buttons offer a simple visual and a fast way to route the conversation as needed . But we won't use these in this project. Remember, we had to images for this purpose to use images as options, I lose the gallery template. Like so, the gallery template, as offered by Facebook, is an excellent way to display products and multiple options, each with its set of buttons for the user to quickly take action. So, for instance, a T shirt store displaying various T shirt designs along with buttons to buy the product in this simple case here will begin by uploading the 07 Iron Man image, and I'll be asked if I want this to be shown as horizontal images or square images. This is depending on the kind of him it's used. We don't need to worry here, so I'll stick to horizontal chat. Fuel also offers this built in cropping tool, which can be used to crop out the image. As for the aspect ratio you've selected here, I can see a horizontal or wide frame, and I'll resize it as follows before hitting the done button there to text fields here. The heading is the main board heading text for the image. So I leave this image Iron Man for the subtitle or description box. I'll dive something like Tony Stark. Yeah, then I'll add one more image to the galley, and we'll upload the Batman image like so Once again, we'll set the heading to Batman and the subheading to something like Bruce Wayne is the best. I'm really trying to awaken the fan in the user. The gallery card, which Facebook actually calls a castle off generate templates, can offer upto 10 options at a time as ah, horizontally scalable list. Now notice that we can also add buttons here. We can add up to three buttons here, but we'll only need one button for each off these image cards. In this project, the Iron Man card will carry a button that says, I'm a Marvel fan and for Batman Will have I'm a D c. Fan button. Depending on what the user clicks, we want to redirect the user toe another block, which offers facts and data about the chosen comic brand. So I'll create two more blocks Marvel Fan and B C fan like so we'll build these out in the next two lectures, so leave them empty for now to make the user go from this block toe. The Marvel fan block, for instance. All you have to do is click on the I'm a Marvel fan button, and from the pop of dialogue, simply select the Marvel fan block. Like so you notice we have a couple of other things that we can do with these buttons for instance, you can link the buttons to an external you earl, or turn it into a phone call button so that a person can dial a given number directly. And you can even add a by feature to allow users to directly purchase products from within . Facebook Messenger Though this is a limited feature available mostly to U. S customers at this time and to customers in a select band, off nations through an integration with stripe anyway, in a similar fashion, will also link up the I'm a D C fan Barton to the BC Fan block. You can see how easy it is to segregate users on the choices they make and send them to domain specific blocks off conversation. But one question still remains how Ruby send the user from the favorite movie block toe. The last question block for this purpose. I'll make use off the go to block plug in that chat feel offers us the go to block. Blufgan is extremely powerful, as you'll see later in this section itself. For now, all we will ask this to do is to redirect the user toe the ask question block simply port when the favorite movie block has played. In the end, it redirects the user to the Ask question block, which starts playing seamlessly and the conversation continues. This branching capability is extremely powerful. As you'll soon discover Let's test this border distinct. I'll type favorite movie, and the favorite movie Block plays out and continues on to the ask question block. As you can see, we can also see our gallery card in action. I can scroll it horizontally, and we have our buttons as well. I won't click these buttons right away, so let's hope on to the next lecture and see what the Marvel Fan Block is all about. 13. BumbleBot Says Hi - Marvel Facts: in this lecture will work on the Model fan block, which is expected to display some facts about Marvel Comics now it would be quite boring to see one factoid every time that we reached this part of the conversation instead will make it more interesting, well equipped. Are bored with three model facts instead of one, and I'm using three. To keep the example simple, you can add as many facts as you want. Then we'll create a mechanism so that a random fact is displayed every time we reach this point in the conversation. Randomly routing the conversation is a powerful feature. Natural human conversation always brings with it a certain degree of chaos, a degree off randomness. With sufficient work, your board can introduce randomness and make the conversation more believable. I'll begin working on the model fan block by first using an image card to bring in an image off Stanley saying Nice suit. After all, how can we even think about discussing model without a guest appearance by Stanley himself ? And I'll also go into the D C fan block? And just like the Marvel fan block, let's create an image card with 09 Joker image like So Okay, Going back to the Marvel fan block, I want to randomly show a fact. But before that, I'll add a two second typing delay, and we'll add a text message saying, Here's an interesting fact about Marvel Comics. At this point, I need a way to branch out toe one off three factoids that I have with me. To accomplish this, I'll first create a new group called Marvel Facts. Here, I'll create three blocks. Every block will have a text card displaying a fact about Marvel Comics. I've included these facts as resource is for this lecture. Another important point that you must keep in mind here is that every block on Chart fuel has a unique name, and nor two blocks can share name. This is irrespective off which group this it within elite. Now let's go back to the model fan block and I'll add Go to block, plug in to randomly route the user toe one off the marvel fact blocks. Didn't I tell you the go to plug in was amazing? All you need to do is enable the random option add the first model fact block, then click on add random block and select the second fact block and so on. You can keep adding blocks, and every time the conversation reaches here it will be randomly and unpredictably routed toe one of the blocks that you've said here. So let's test this by saying your favorite movie. And when we reach this point, I'll click on I'm a model fan and, as expected, I'm presented with the fact at the end, we'll also create an A rule to see the facts directly without needing to go. Why are the favorite movie block for this purpose? I create a new block in the model fact section, and we'll call this show a marvel fact. All we need here is a go to plug in that randomly routes the users toe one off the three Marvel fact blocks like So then we can create an A rule for phrases such as Give me a marvel fact. Show me a model fact, tell me about Marvel Comics, etcetera, and they should directly in walked the show a Marvel fact block listlessness. I'll type Show me a marvel fact and I get a fact right away. I loud type one model fact please. And I get the fact again it should be noted that the outcome off the randomness can never be predicted. So with fewer options to randomly route toe, you might end up on the same fact again in successive tries before it actually shows you are different and a random fact here. After three tries, I finally see your and in fact, to prevent this just Ardmore facts. More branches, all right. With this done, let's also work on facts for DC Comics fans. 14. HelloBot - DC Facts: all right. You know the drill. Just like we build the facts for Marvel fans. Let's now build one for BC fans. And because the process is exactly the same instead, off me talking over the lecture, I'll simply ask you to watch this lecture while I leave you with Delvin Lamar. Organ trials between the mustard and mail. Okay, No. - Okay , I'm interrupting the music to discuss one important point here since we made the show a model fact block which is tasked with routing the conversation randomly toe one of the three fact blocks. I don't need to specifically create a randomly setting goto block in the Marvel Fans sequence, as you can see here. Instead, we will simply send the conversation over to the show a model fact block which does the random groping for us. 15. Handling Small Talk: So now that are hello board has done. We should also spend a while in equipped the board to handle small talk. Small dog refers to random conversation, which is not usually domain specific. Such messages may or may not be important, and the board should be designed to handle some, if not all, of these messages to begin with. For instance, the user might just say hi hello by catcher later. Who are you? Help me and several other phrases that may or may not fall under a very specific category the baby can handle. Small talk on Jack Fuel is by creating a rules. So let's get back to our board and add some rules to handle small talk. I'll go into the set up a section once again, and what I will first do is click here and rename this group of rules that we have been sitting in this project to bought rules. I'll then create a new group called Small Talk, which will house all our small talk rules. Groups in this section allow you to organize rules in a fantastic week, and another reason why you would do that is because you can export these groups off rules from one board toe, another in your account. This is an incredible feature because you don't have to recreate a common set of rules Every time you build a chat board for a client board specific rules and common rules, such a small talk can be kept separate. Our first rule addresses common greetings such as High Hello, Hey, etcetera. And instead of a single fixed response, we can easily randomize from a set of text responses like So I have not created a block in this case because it is a simple greeting and nothing more and a simple text response will do. I should also point out that the exact nature off responses would depend on the board your building, For instance, you might want to include helpful instructions in these messages to prevent the user from drifting away. Okay, lets test this first rule. You can see at the invocation. It randomly throws a unique response from the given set. The next rule is when a user might ask, What do you do, or who are you? Or help me and so on? In this case, you can include a more useful set off responses like so next up. When the user says by CIA catch you later, we can do the same. Finally, when someone asks, Are you really or are you abort? It never hurts to be ready. This process off training the board to respond to small talk as well as other rules, is a continous one and must keep going, especially when the boat goes live. There are analytics on chat fuel that we will see later that can show you what the users are typing in when the chat with your board. And based on that, you can iterative Lee ad rules and improve how your board response. So in actual you'll have toe continually make your board smarter. So there you have it. Your very first chart board on chat fuel is done. Let's summarize in the next lecture. 16. Summary: congratulations on building your first chat board. We created a simple board that just answers to basic questions. The idea was to introduce you to the fundamentals of jack fuel by building an actual chair port. We began by connecting our board to a Facebook page. Then we learned about the role. Off blocks and cards. Blocks are containers that let you organize a sequence of cards in the context, off a conversation or a sequence off messages. We used cards as building blocks. Guards individually represent text messages, images, gallery off images and text, audio video and many more types, off messages and media. Chad fuel allows you to easily add various types of media toe visually, let you build a sequence of messages. We also understood the rule off the welcome message block, which is played out when a user first interacts with the body. Likewise, we saw the default answer block, which plays out when the borders unable to understand user input, we saw how groups can be used to organize these blocks as well. Now, when working with guards, we saw the use off system attributes which are read only properties that bring in data from Facebook, such as the users first and last name, among other things. These can be used to personalize responses to the user when displaying text cards. In particular, we used the DeLay plug in to introduce simulator typing. Deal is this makes the conversation sequence more interesting As it's similar. It's the delay that a real person would introduce while typing out messages. This makes the conversation more believable and natural, and also gives time for the user to read and understand messages that the boat sends. We also saw the amazing go to Plug in, which enables us to route the user from one block to another. And we also saw how this plug in enables us to randomly route the user toe a block from a given set. The Goto plug in is a real time saver. When creating complex board flows, Chatfield also offers a built in natural language understanding system that enables us to identify what the user is saying and perform in action. These are called a rules, and we saw how these rules can be set up and used to route conversation, and also for handling simple small talk. So there you have it. If you have any queries, feel free to interact with me on the forums or send me a message 17. Congratulations: congratulations on making it to the end off this class and on building your very first chat board. Though our chapel was pretty basic, it introduced you to the process of building conversational us and using Chatfield toe build and deploy a chat port on the Facebook Messenger platform. But we're not done yet. We're just getting warmed up. Watch out for the next class in the chat board boot camp series where we take things a notch higher and create a practically usable chap work. All of this is coming soon on sculpture under Len. Thank you.