1 00:00:00,007 --> 00:00:02,690 - [Instructor] In this and the next several videos, 2 00:00:02,690 --> 00:00:04,510 I want to take a little bit of time 3 00:00:04,510 --> 00:00:08,080 to overview some of Watson's services. 4 00:00:08,080 --> 00:00:09,930 Now, this may seem a little bit 5 00:00:09,930 --> 00:00:11,730 like a marketing presentation. 6 00:00:11,730 --> 00:00:14,120 But the main reason that I want to do this 7 00:00:14,120 --> 00:00:18,336 is to tie together these kind of pre-bundled capabilities 8 00:00:18,336 --> 00:00:21,700 that you can incorporate in your own applications 9 00:00:21,700 --> 00:00:25,270 with some of the concepts that we presented 10 00:00:25,270 --> 00:00:27,309 not only in the last two lessons 11 00:00:27,309 --> 00:00:30,470 on natural language processing and data mining Twitter, 12 00:00:30,470 --> 00:00:33,300 but also on the subsequent lessons where we're going 13 00:00:33,300 --> 00:00:35,704 to get into machine learning techniques, 14 00:00:35,704 --> 00:00:37,920 deep learning techniques, 15 00:00:37,920 --> 00:00:42,630 and big data and IoT techniques as well. 16 00:00:42,630 --> 00:00:45,670 So not only do you have pre-bundled capabilities 17 00:00:45,670 --> 00:00:48,850 from Watson and other similar cloud vendors, 18 00:00:48,850 --> 00:00:51,470 like Amazon, and Google and Microsoft, 19 00:00:51,470 --> 00:00:53,773 but you also have the ability to 20 00:00:53,773 --> 00:00:57,825 take it back a level and build your own models 21 00:00:57,825 --> 00:01:00,590 for a lot of capabilities as well. 22 00:01:00,590 --> 00:01:02,330 And that's what we'll start showing you 23 00:01:02,330 --> 00:01:04,150 in the next couple of lessons. 24 00:01:04,150 --> 00:01:06,090 So to start out, I want to talk briefly 25 00:01:06,090 --> 00:01:07,530 about the Watson assistant, 26 00:01:07,530 --> 00:01:10,520 which helps you build virtual assistants or chatbots. 27 00:01:10,520 --> 00:01:12,409 Many of you have probably used these 28 00:01:12,409 --> 00:01:14,860 in various websites nowadays. 29 00:01:14,860 --> 00:01:18,610 In fact, the first line of customer service nowadays 30 00:01:18,610 --> 00:01:21,154 is often a chatbot to get you through 31 00:01:21,154 --> 00:01:23,800 to the correct person or department. 32 00:01:23,800 --> 00:01:26,300 And one of the things that's nice about Watson assistant 33 00:01:26,300 --> 00:01:29,000 is they have built in what they call 34 00:01:29,000 --> 00:01:32,840 intense for many common scenarios already. 35 00:01:32,840 --> 00:01:35,550 And this link is going to show you how to take 36 00:01:35,550 --> 00:01:39,300 advantage of their built-in assistant catalog. 37 00:01:39,300 --> 00:01:41,210 But separately, you have the ability 38 00:01:41,210 --> 00:01:45,510 to train Watson assistant for your own scenarios 39 00:01:45,510 --> 00:01:47,567 for whatever application you're building. 40 00:01:47,567 --> 00:01:50,100 So if you were building a weather chatbot, 41 00:01:50,100 --> 00:01:54,070 you could enable it to respond to natural language questions 42 00:01:54,070 --> 00:01:56,740 like "what is the weather forecast for New York City?" 43 00:01:56,740 --> 00:02:00,010 Or if you're building a customer service chatbot, 44 00:02:00,010 --> 00:02:03,900 it could answer questions specific to your company 45 00:02:03,900 --> 00:02:06,730 or product. But it also might have the ability 46 00:02:06,730 --> 00:02:09,710 if it can't answer something to direct the person 47 00:02:09,710 --> 00:02:12,430 through to somebody who can help them perhaps 48 00:02:12,430 --> 00:02:14,890 a live person in that case. 49 00:02:14,890 --> 00:02:17,740 And most interestingly, perhaps is this 50 00:02:17,740 --> 00:02:19,740 second to last link on the slide, 51 00:02:19,740 --> 00:02:22,703 which is the IBM code bot exchange. 52 00:02:22,703 --> 00:02:26,420 On this website, you can see lots of 53 00:02:26,420 --> 00:02:29,250 pre-existing bots and virtual assistants 54 00:02:29,250 --> 00:02:30,900 that other people have built. 55 00:02:30,900 --> 00:02:32,850 And if you build your own, you can also 56 00:02:32,850 --> 00:02:35,470 share yours there as well. 57 00:02:35,470 --> 00:02:38,330 So the demo for this service is at the page 58 00:02:38,330 --> 00:02:40,080 that you see at the bottom of the screen. 59 00:02:40,080 --> 00:02:41,940 Let me go ahead and pull that up here. 60 00:02:41,940 --> 00:02:44,340 There's actually three demos that they have 61 00:02:44,340 --> 00:02:47,070 of different chatbots scenarios here. 62 00:02:47,070 --> 00:02:50,880 If you go ahead down to the bottom left corner here, 63 00:02:50,880 --> 00:02:53,140 you'll see this is the make a payment demo. 64 00:02:53,140 --> 00:02:55,278 And the arrows will allow you to navigate 65 00:02:55,278 --> 00:02:57,561 through the other demos as well. 66 00:02:57,561 --> 00:02:59,690 So the assumption here is that you need 67 00:02:59,690 --> 00:03:01,300 to make a credit card payment 68 00:03:01,300 --> 00:03:05,229 and that you've logged into your bank's server 69 00:03:05,229 --> 00:03:06,570 for your account. 70 00:03:06,570 --> 00:03:10,530 It then recognizes what your cards are with that bank. 71 00:03:10,530 --> 00:03:13,490 And you have the ability not only to click buttons 72 00:03:13,490 --> 00:03:14,730 representing your accounts, 73 00:03:14,730 --> 00:03:16,950 but you can also type down here as well. 74 00:03:16,950 --> 00:03:19,330 So let's say I want to pay for this card, 75 00:03:19,330 --> 00:03:22,070 it's telling me when the payment is due, 76 00:03:22,070 --> 00:03:25,240 I can either choose to pay today or on a given date. 77 00:03:25,240 --> 00:03:26,960 So let's say I want to pay today. 78 00:03:26,960 --> 00:03:29,040 It tells me what my minimum balance is 79 00:03:29,040 --> 00:03:31,020 and what my total balance is. 80 00:03:31,020 --> 00:03:34,510 So let's say I decided I wanted to pay $25. 81 00:03:34,510 --> 00:03:37,700 It is going to know that that's not the minimum 82 00:03:37,700 --> 00:03:39,524 and it's going to tell me that I have to 83 00:03:39,524 --> 00:03:41,960 enter another value here. 84 00:03:41,960 --> 00:03:44,130 So let's say I want to do the minimum value. 85 00:03:44,130 --> 00:03:47,520 And now it's going to go ahead and issue that payment 86 00:03:47,520 --> 00:03:48,880 once I confirm it. 87 00:03:48,880 --> 00:03:52,060 So as you see here, I'm interacting with natural language, 88 00:03:52,060 --> 00:03:55,010 and now it shows me my updated balance. 89 00:03:55,010 --> 00:03:58,430 So just a simple chatbot experience 90 00:03:58,430 --> 00:04:01,393 but something that you can build very easily with Watson.