Building Chatbots with Amazon Lex and IBM Watson

Build a subway bot that orders subs using Amazon Lex and IBM Watson, testing intents, bread options like Italian, and bot responses through an integration.

Connect to your AWS account, sign in as a new user to create your profile, and explore region us-east-1 with Lex and Lambda to build chatbots.

Explain the Amazon Lex interface, create and customize a bot with intents and slot types, test samples, configure error handling and analytics, and begin building a sub-ordering conversation.

Create and configure intents by selecting existing or creating new ones, define utterances and slots, set up validation and prompts, and enable fulfillment via a Lambda function or API.

Create and organize Lex slot types using a Subway-inspired naming convention, defining bread type, size, sauce, toppings, and temperature values; align with utterances and prepare for tests.

Learn to set up slots and utterances in the intents view, renaming utterances to bread, size, and sauce, and configure prompts and priority to ensure required parameters are collected.

Learn to implement error handling and user clarification prompts in Amazon Lex, test and train the bot, and integrate it with Twilio for SMS conversations.

Test the bot by demonstrating a confirmation prompt and required slot filling for a sub order, wiring Amazon Lex to Lambda for fulfillment, and preparing for a Twilio channel.

Create a Twilio account and explore trial mode messaging, learn about paid plans for production, and locate your account summary and auth token to enable Amazon Lex integration.

Configure Twilio on AWS with webhooks by creating an alias for subway orders, entering the authentication token, activating the integration, and testing the endpoint in the Twilio console.

Set up a Twilio messaging service to enable a two-way interactive chat bot, configure inbound callbacks, attach a phone number, and prepare for testing on the phone.

Test the Twilio and Amazon Lex integration by texting a sub order, confirming bread type and the disconfirmation prompt, and validating error handling and intent flow.

Advance your practice session by building a burger ordering bot in Lex, defining an ordering intent with buns, toppings, sauces as slots and utterances, and integrating with Twilio to text.

Learn to build a hotel booking chatbot with Amazon Lex that collects required data through prompts, and explore Facebook integration with Facebook cards.

Build a go book a room intent in AWS Lex by defining slot types for number of people, location, and room options, and configure utterances and error handling with retries.

Create utterances to let users invoke the hotel booking chatbot with simple phrases such as hi or I want to book a room, and collect details like location and rooms.

Create and organize slot types for a Lex chatbot by renaming and adding location, time, and room slots, using Amazon time as a default, and presenting options with Facebook cards.

Set up Facebook integration for Amazon Lex by modifying Amazon settings, creating a Facebook app and page, and configuring tokens and webhooks so Lex and Facebook can communicate.

Create a Facebook page and obtain a page access token and app secret by configuring developer settings, creating a Facebook app, and enabling Messenger for Amazon Lex integration.

Learn how to integrate Amazon Lex with Facebook Messenger by configuring a callback, verify token, page access token, secret key, and token generation, then test and publish.

Publish the Lex-powered booking bot to beta and test Facebook messenger with cards showing city options like San Francisco, Palo Alto, and San Jose, including check-in times.

Create a flight-booking bot for Facebook Messenger by defining slots (classes, duration, location, costs), building a new bot and intents, and testing utterances.

Create a new Lex bot named recruiter service, set the Joana voice, add the first intent, and outline simple error handling and upcoming slot types in the next video.

Explore how to create utterances and slot types in Lex, configure job-search prompts, collect details like location, citizenship, and years of experience, and prepare job listings for Slack integration.

test the Amazon Lex bot by building and testing the app, observe responses, fill slots, and prep Slack integration by creating accounts, a group, and configuring the API.

Set up a Slack workspace, create an alias, verify email, and explore channels and direct messages; prepare to connect the Slack API and obtain the client secret and client id.

Set up a Slack app and bot user, enable interactive messages, and record the client secret and verification token to enable conversation between Amazon Lex and Slack.

Connect an Amazon Lex bot to Slack by wiring postback events, assigning permissions for posting and reading, subscribing to Slack events, and testing the integration.

Test the Slack API integration for recruiter interactions, including user authorization, direct messaging, and conversational prompts about job details and experience.

Explore IBM Watson's conversation tool to build chatbots that understand natural language, connect across channels like Slack and Facebook Messenger, and scale with enterprise-grade infrastructure.

Explore how IBM Watson uses intents, greetings, and entities to drive dialogue flows, create greeting responses, and test bot interactions in the testing space.

Set up an IBM Conversation account and learn to build and deploy bots across channels using IBM Watson, within a 30-day free trial.

Begin by selecting the Watson Conversation service in the IBM DMX catalog, create an instance in a nearby region, and review the light 30-day plan and API pricing.

Create a new workspace and define intents and entities to build dialogue with IBM conversation. Test conversations, monitor analytics, and deploy to Slack.

Create and organize IBM conversation intents for a customer service bot, including greeting, goodbye, complaint, and return, with examples and anything else, in the next video for entity modeling.

Create and configure IBM conversation entities by defining values and synonyms, such as return items, books, and birds, to map user input to accurate chatbot responses.

Create and configure Watson conversation dialogues, starting with greeting and complaint nodes, handle unrecognized input via intents and entities, and design returns with child nodes and specific entities.

Explore IBM conversation dialog concepts by creating child return nodes with conditions, setting parroted responses for specific issues, and linking jumps to guide testing and graceful fallbacks.

Test your Amazon Lex and IBM Watson chatbot by using the test icon to enter keywords, validate greetings and complaints, inspect intents and entities, and refine responses before ending conversations.

Modify the return variables to include computers, tablets, and phones. Add these items to the dialogue and entity screen, and test in the sample test dialog.

Build a restaurant reservation chatbot with IBM Watson, detailing opening hours, location, and table bookings, and test multiple use cases in the IBM conversation testing simulator.

Create intents in IBM conversation to build a restaurant booking service, covering reservation, change, cancel, confirm, hours, location, and greetings to train natural language understanding.

Create entities for date, guest number, and time, with day names and synonyms (today, tomorrow, next Friday) to drive tailored responses, then prepare for dialogue and messenger calendar integration.

Learn to design a restaurant reservation dialogue using intents, entities, and context, starting with a welcome, handling hours and location, and building a reserve table flow with JSON-style code.

Build a restaurant reservation dialog by validating guest number, date, and time, then use conditional blocks and node jumps to confirm or change reservations.

Create a layered restaurant reservation dialog using json-based nodes for date, time, and guest number, building and validating a multi-parameter flow with abort and off-topic handling.

Create a restaurant reservation dialog by capturing date, time, and guest number with structured variables, abort options, and conditional jumps to confirm and finalize the booking.

Test the restaurant reservation chat flow by validating hours prompts (3 to 11:30 p.m.), then book a table for tomorrow at 9 p.m. or 1:00 p.m., and note improvements.

Practice problem for the restaurant reservation bot adds lunch and dinner intents with lunch options and synonyms, connecting utterances and entities to extend the existing codebase.