This is the cloud computing asssignment. AI Customer Service Assignment
Implement the five components we discussed in the lectures: API Layer Services: API Gateway, Lambda, Cognito, IAM Requirements: Create a Cognito User Pool Create a Cognito Identity Provider for the User Pool above
Implement the following APIs: /context GET Retrieves all the context events for a particular user /biometrics PUT Communicate with the biometric layer and save a new biometric photo /biometrics/verify POST Given an incoming image, verify that the image matches the user Enable IAM security on the three API Gateway APIs above Modify the Authenticated User IAM role that Cognito creates to allow the role to call the API calls above
Context Layer:
Services: Kinesis, Lambda, DynamoDB Requirements: Given an event sent to Kinesis, trigger a Lambda function, validate and augment the event as necessary and then store it in a DynamoDB table that hosts context Use DynamoDB’s expiration feature to expire context events after 15 minutes.
Biometric Layer:
Services: Lambda, Rekognition, S3, DynamoDB Requirements: Create two Lambda functions that Given an incoming photo, store it to a specific biometric pool of photos in S3 and store a reference to that S3 file in DynamoDB Given an incoming photo, compare it to a specific biometric pool of photos (references stored in DynamoDB) using Rekognition and return if it is a match or not
Chat Layer:
Services: ElasticBeanstalk, Lex, Lambda, S3 Requirements: Deploy the Node.js server to ElasticBeanstalk Use Lex to disambiguate the user utterances Add at least 3 intents Use Lambda to elicit the Lex slots Create a webhook (an endpoint) to receive sentiment analysis results and send them either in an email using SES or back through the chat to the user At the end of a conversation store a JSON file in S3 with context surrounding the conversation (conversation id, user id etc.) and the user utterances
Analytics Layer:
Services: S3, Lambda Requirements: Given the JSON object stored in S3, trigger a Lambda function that iterates over the user utterances and outputs an overall sentiment for the conversation Send the sentiment analysis back to ElasticBeanstalk to a webhook Download the sample iOS application showcased during class and provide the required information to test your components.