AWS is a leader with regards to services provided in the cloud. Learning all of them is a time-consuming and tedious exercise. I put together a brief overview on the AWS AI Services available today.
I had excluded all marketing content and concentrated on providing only a short description, in the end I do not have to sell these services :). My goal is to help a machine learning consultants to get a quick feel of the AWS AI Services functionality. You will find reference link to every service. Depending on your project at hand you can dive deep into a specific one.
AWS Machine Learning Specialty
This post is also part of a series about the AWS Machine Learning Speciality (MLS-C01) exam. I have structured the content into seven separate posts. These posts can be consumed as a stand-alone material. If you are not preparing for the MLS-C01 exam you may still find the topics interesting:
- Part 1: The AWS Certified Machine Learning Consultant
- Part 2: The AWS Data Engineering Consultant
- Part 3: Top 7 Explorative Data Analysis Methods
- Part 4: Machine Learning Algorithms Overview
- Part 5: AWS AI Services Overview
This post covers one of the domains in the MLS-C01 exam. Domain 4. Machine Learning Implementation and Operations:
- 4.1 Build machine learning solutions for performance, availability, scalability, resiliency, and fault tolerance.
- 4.2 Recommend and implement the appropriate machine learning services and features for a given problem.
- 4.3 Apply basic AWS security practices to machine learning solutions.
- 4.4 Deploy and operationalize machine learning solutions.
AWS AI Services
The main page of the AWS AI Services can be found here: https://aws.amazon.com/machine-learning/ai-services/
Amazon Augmented AI
Amazon Augmented AI – helps to integrate human review of low confidence machine learning predictions. The required the following artifacts to be set up:
- Work force: group of people from Amazon Machanical Turk, 3rd party vendor or internal group of people.
- Worker task template: web UI layout about your task.
- Flow definition: this workflow connects the template and work force using a set of steps to be executed. It integrates with other AWS services via build-in task types (Textract, Rekognition).
- Human loop: effectively start the defined workflow for a given data object.
Amazon CodeGuru
Amazon CodeGuru Reviewer – helps to find performance problems based on Java source couse stored in GitHub, Bitbucket or AWS Code Commit.
Amazon CodeGuru Profiler – collects runtime performance data from JVM based live applications.
Amazon Comprehend
Amazon Comprehend – helps to analyze natural text documents with many features: Keyphrase Extraction, Sentiment Analysis, Syntax Analysis, Entity Recognition, Comprehend Medical, Custom Entities, Language Detection, Custom Classification, Topic Modeling, Multiple language support, Medical Ontology Linking. API reference.
Amazon Forecast
Amazon Forecast – helps to create time-series forecasting applications by using provided algorithms and parameters (data frequency, forecast horizon, etc)
Amazon Fraud Detector
Amazon Fraud Detector – helps to identify potentially fraudulent online activities such as online payment fraud and the creation of fake accounts. It supports model training on reference data and rule creation.
Amazon Kendra
Amazon Kendra – helps to create enterprise search service powered by machine learning.
It integrates with: S3, SharePoint, Salesforce, Servicenow, RDS databases, One Drive.
Index creation and search are the basic operations.
Amazon Lex
Amazon Lex – Alexa like conversational interface, use case example: chat bot.
Input: Automatic Speech Recognition (ASR) and Natural Language Understanding (NLU).
- bot – high level object to represent a use case,
- intent – action configuration:
- sample utterances – examples sentences to trigger the intent
- slot – intent parameters (number, text, phone number, email …)
- action – lambda function to execute when user specified all input
Amazon Personalize
Amazon Personalize – helps to implement real-time personalization and recommendation functionality. “It combines user interaction data with contextual data to generate high-quality recommendations.”
Basic usage flow:
- import user and product data,
- create a solution (trained model),
- create a campaign (from a deployed solution version),
- get recommendations from the campaign campaign id using API by providing user id or product id.
Amazon Polly
Amazon Polly – helps to turn text into speech, using different languages, dialects and genders. It supports Speech Synthesis Markup Language – SSML, to fine tune the speech (breaking, emphasizing, whispering …)
Amazon Rekognition
Amazon Rekognition – helps to analyze image and video files. Image features: Object And Scene Detection, Facial Recognition, Facial Analysis, Face Comparison, Unsafe Image Detection, Celebrity Recognition, Text In Image.
Video features: Real-Time Analysis Of Streaming Video, Person Identification and Pathing, Face Recognition, Facial Analysis, Objects, Scenes And Activities Detection, Inappropriate Video Detection, Celebrity Recognition
Amazon Textract
Amazon Textract – helps to extract text from scanned documents using simple request response API.
Amazon Transcribe
Amazon Transcribe – helps to convert speech (audio file or stream) from major languages to text file, features: timestamp generation, multiple speakers and channel identification. It works with a simple request response API.
Amazon Translate
Amazon Translate – helps to translate texts to other (55) languages in real time or async. It works with a simple request response API.
AWS DeepLens
AWS DeepLens – Wi-Fi enabled digital camera sold by Amazon which can integrate to AWS services for real time model evaluation using AWS Lambda (via AWS IoT Greengrass)
AWS DeepRacer
AWS DeepRacer – Wi-Fi enabled remote control car to be trained with reinforcement learning.
AWS DeepComposer
AWS DeepComposer – AI enhanced music keyboard.
Tipp: practically for the exam you should be able to chain these models as required, e.g.: voice translator, healthcare chatbot.
Conclusion
This post was a dense summary of the AWS AI services. In the previous posts I have introduced the AWS Machine Learning Speciality certificate for AWS Certified Machine Learning Consultants, AWS Data Engineering services for The AWS Data Engineering Consultants, the Top 7 Explorative Data Analysis Methods and Machine Learning Algorithms Overview. In the next post I will present Amazon SageMaker. See you there!
Photo by Paul Hanaoka on Unsplash