Building Apps Using Amazon's Alexa and Lex

Explore how Alexa enables voice interactions, examine voice interfaces and Alexa skills basics, including invocation, utterances, and intents, and learn AWS integration and testing tools.

Trace the evolution of human-computer interaction from punched cards to voice interfaces, and explain how voice recognition and NLP power smart assistants and IoT commands.

Unlock the Alexa ecosystem across Echo devices and AWS cloud by learning to build voice experiences with the Alexa Skills Kit and Alexa Voice Service for Amazon and third-party devices.

Learn how invocations, utterances, and intents drive Alexa skills in the cloud, with the Alexa service routing voice requests from Echo to AWS-hosted skill APIs.

Explore AWS Lambda and DynamoDB to host code and store data, configure Alexa skills in the Amazon developer console, and test end-to-end with Echo sim.io, using the free tier.

Explore the four kinds of Alexa skills—custom, smart home, flash briefing, and video—and how utterances map to intents and content is delivered.

Build a stock market tracker Alexa skill using finance APIs and Lambda functions. Learn how utterances map to intents via the Alexa service to report the Dow Jones Index.

Explore how Alexa maps utterances and invocations to intents and how a custom skill can handle multiple intents with many utterances, using slots to fill in dynamic values.

Fetch real-time and daily Dow Jones data from Alpha Vantage with a free API key, and configure intraday and daily time series to access open, high, low, and close values.

Configure your first Alexa skill by defining the interaction model, intents, and sample utterances in the Alexa Skills Kit, then connect AWS Lambda via ARN and test end-to-end.

Discover how AWS Lambda provides a serverless, pay-as-you-go compute model with automatic scaling, and how to connect it to an Alexa skill via ARN to fetch live Dow Jones data.

Create a lambda function, assign a role and permissions, and set the Alexa Skills Kit as the trigger; connect the skill to the lambda using the lambda URL and ARN.

Test your stock market tracker lambda code iteratively to ensure Alexa integration. Use unit tests in the lambda console, Alexa skills kit utterances, and end-to-end echo.io tests.

Test the Alexa skill launch by returning a json response with speech and a stock tracker card, using helpers to build the speech and set shouldEndSession to false.

Code and test an Alexa custom intent 'get index' by calling Alpha Vantage to fetch the latest Dow Jones close, parse JSON, and build a speech response with a card.

Learn to debug Alexa skill lambda code by testing iteratively, add and handle the Amazon help and stop intents in the interaction model, and use CloudWatch logs for troubleshooting.

Practice end-to-end testing of an Alexa skill using echo.io, including activating the microphone, observing delayed voice command processing, and reviewing market tracker responses and Alexa app cards.

Explore how slots function as request configuration parameters to customize Alexa stock tracker dialogues, enabling multi-turn conversations that collect stock name, date, and price type.

Explore the Alexa dialogue model by defining slots and prompts for stock, price type, and date, enabling dialogue to fill all slots and fulfill requests.

Access stock data with the Ingenio API by signing up, retrieving an access key, and making authenticated requests to obtain Apple’s closing price on 15th September 2017.

Define slots and their values with synonyms to make utterances flexible, then build the stock tracker dialogue skill using three slots (stocks, price type, date) in the Alexa skill builder.

Configure Alexa slots for the get stock info intent using user defined and Amazon date slots, with prompts and sample utterances to collect price type, stock name, and date.

Link your Alexa skill to an AWS Lambda function, handle session start, launch, and end requests with a JavaScript handler, and use reprompt and welcome responses to guide users.

Demonstrate handling the get stock info intent, filling stock name, price type, and date slots with dialogue delegate prompts, and fetching stock prices from a backend using basic authentication.

Demonstrate handling help, stop, and cancel intents in an Alexa skill by mapping help to the welcome response and stop or cancel to the goodbye response in lambda.

Test stock tracker dialogs with echosim.io to retrieve opening, closing, high, and low prices for IBM and Amazon on October 4th, 2017 and October 2nd, 2017.

Learn to persist user data across sessions with DynamoDB, a NoSQL database on AWS, enabling a stock list skill to remember favorites via key-value storage.

Instantiate a DynamoDB on Amazon Web Services to create a stock list table with user ID as the partition key, ready for use in your Lambda and Alexa skill.

Configure a Lambda function with a DynamoDB full-access role to persist a stock list across sessions in an Alexa skill, covering IAM roles, Lambda, and DynamoDB integration.

Learn to implement Alexa skill handlers in Lambda, including session start, launch, intent, and session end requests, and build responses with a welcome message and add, remove, or list stocks.

Handle add, remove, and list stock intents for an Alexa skill backed by DynamoDB, using AWS SDK document client to update and get the stocks list by user ID.

Test voice commands with echosim.io to add, remove, and list stocks in a stock list, with IBM, Microsoft, and Amazon as favorites.

Learn how to build a flash briefing skill for Alexa that delivers short text or audio content from RSS feeds, with content you specify and utterances defined by Alexa.

Set up a Twitter account and compose tweets to feed into an RSS feed via Twitter RSS, then fetch the RSS URL for use in a flash briefing.

Configure a flash briefing skill named Market Flash in the Alexa developer console, using the twit RSS URL as a text feed with a preamble.

Learn to design and test Lex chat bots and Alexa voice interactions, mastering intents, utterances, slots, and lambda-based fulfillment. Build and deploy bots with Slack, Facebook, and Twilio integrations.

Sign in to the AWS console and explore Lex in the AI services. Learn how Lex supports voice and text chat, with intents, slots, and Polly speech.

Compare custom bots with sample bots to understand how intents, utterances, slots, prompts, and fulfillment shape conversations, and explore templates for booking trips, ordering flowers, and scheduling appointments.

Explore how Lex bots use IAM roles to securely access AWS services such as Polly, Lambda, and CloudWatch, enabling text and voice interactions with session context and fulfillment logic.

Create a Lex chat bot by outlining intents and slots in the editor, and configuring channels, monitoring, and settings. Learn which channels Lex supports—Facebook, Slack, and Twilio—and COPPA considerations.

Explore building Lex chatbots by defining intents and slots, including built-in slots and sample utterances, test in the Lex console, and handle confirmations and errors.

Discover how to create and configure intents in Amazon Lex, differentiate intents from user utterances, and map utterances to fulfillment logic using AWS Lambda and built-in intents.

Explain slots as parameters for intents and how Lex uses built-in and custom slot types to map user utterances to values, using a crusts slot example with synonyms and resolution.

Learn how Lex parses utterances into intents using slots, with two slot resolution options: expand values or exact values with synonyms, plus how prompts and retries shape slot elicitation.

Learn how Amazon Lex dialogs use slots, slot types, and priorities to map required values to intents like order pizza and order coke, with cross-bot reuse.

Discover how sample utterances drive a Lex bot by training the underlying model, mapping input to slots like size, crust, and pizza kind, and routing to the lambda function.

Read back the user's order with a confirmation prompt to ensure accuracy, while Lex uses initialization, validation, and fulfillment via AWS Lambda or HTTP endpoints.

Explore how a Lex bot handles gibberish and nonsensical input by configuring bot-level error handling, clarification prompts, and maximum retries to prevent unwanted fulfillment.

Build and test a Lex bot, master intents, slots, and dialog state, and publish the bot for production across Slack, Facebook, and mobile apps.

Explore how Lex uses a dialog state to guide a pizza order, prompting for crust, pizza kind, and size, and how the state becomes failed if confirmation is declined.

Explore how Lex uses a state machine to manage pizza orders, showing how upfront slot values can skip illicit slot and illicit intent states and move to confirm and fulfillment.

Follow a top-down deletion of Lex resources: delete the bot, then the intents, then the slot types, to free AWS resources and avoid unexpected charges.

Implement lambda functions as the fulfillment for Lex bots by building a weather bot that queries the open weather API to return city temperatures, using python 2.7.

Create a weather bot that prompts for location to fetch temperature via a third party API through AWS Lambda, using built-in slot types and reusing intents across Lex bots.

Discover how to use built-in slot types in Amazon Lex to train a bot, extract a location for weather requests, and recognize cities and other entities with prompts and utterances.

Learn to set up a weather bot in Amazon Lex, including slots, confirmation prompts, and cancel dialogs, test the model, and implement Lambda fulfillment with intent reuse across bots.

Explore how AWS Lambda powers Lex bot fulfillment with auto scaling, using stateless code and pay-per-use pricing. Learn supported languages and how Lambda integrates with SNS and Kinesis.

Explore how blueprints deliver sample configurations that wire event sources to Lambda functions for Lex bot intents, and author from scratch versus templates in a Lex-Lambda workflow.

Explore how a lambda handler works with a dispatch function to route Lex bot intents, such as get today's temp, calling get weather and close to return a response.

Create a Lambda handler that configures logging with a logger, sets the time zone, and dispatches intents to a get weather function for Lex, while emitting CloudWatch logs for monitoring.

Learn to extract slot values from the fulfillment lambda by querying the intent request's slots for the location, then build a weather response via session attributes and fulfilled close method.

Configure an AWS lambda function by setting the lambda handler and IAM role, then test a Lex test event to ensure correct invocation and fulfillment.

Learn two Lex fulfillment options—an AWS Lambda function or returning parameters to the client—by wiring the weather bot's get today's temp intent to Lambda and testing.

Invoke the open weather Json API via an AWS Lambda to fetch current weather by city name, then parse and return the result to the chatbot.

Learn to invoke the OpenWeather API from an AWS Lambda function by switching from inline editing to a zip upload, using the requests library and secure API keys.

Package your python lambda with its source code and all third-party libraries in a zip file. Upload to aws lambda and set the handler to main.lambda_handler.

Publish a Lex bot by creating aliases that point to specific versions, enabling seamless client upgrades and supporting development, beta, and production workflows.

Deploy a published Lex bot to Slack, create a Slack app with a bot user, wire OAuth and postback URLs, and enable weather lookups via AWS Lambda and OpenWeather APIs.

Learn how to integrate a weather bot built with Lex into Slack by creating a Slack app, adding a bot user, and enabling interactive messages.

Enable Lex integration with Slack by configuring the postback url as the endpoint for messages, generating OAuth and postback urls, and securely exchanging Slack credentials via a key management system.

Configure Slack integration for a Lex bot by using OAuth authentication, copying the postback and OAuth URLs into Slack, and set bot permissions and event subscriptions for all messages.

Learn how to publish and test a weather bot in Slack by using the manage distribution tab, adding to Slack, authorizing access, and validating responses across London, Paris, and Bangalore.