DEV Community

Cover image for AWS Machine Learning Tools in 2021
K for Dashbird

Posted on

AWS Machine Learning Tools in 2021

This article was originally published on https://dashbird.io/blog/aws-machine-learning-tools-2021-edition/

When you want to stay ahead and on top of things in a fast-moving industry, machine learning (ML) is surely one of the trending solutions. Today, innovative companies already have leading Machine Learning tools well-integrated into their processes. In comparison, your start could seem dreadfully slow. Or maybe you just don't have the time or resources to invest in running your own Machine Learning training infrastructure.

AWS offers over 20 services alone in its machine learning category, and that's not counting the other services that have soft Machine Learning features integrated. The services range from low-level offerings like SageMaker, which helps build and manage infrastructure for your learning environments, to high-level systems like Rekognition that come with pre-built Machine Learning models for image recognition.

So depending on your problem, you can start training your own models on AWS' infrastructure or simply pump your data into pre-trained models to extract additional value from it.

This article will go over the current Machine Learning services AWS has to offer.

The Base of Machine Learning with AWS

While you can build your Machine Learning infrastructure by hand from containers or virtual machines, Amazon SageMaker forms the base of all the Machine Learning tools AWS has to offer. It is a service that streamlines all the Machine Learning tasks that come up, from preparing data and building a model to training, and deploying it. SageMaker is the first integrated cloud IDE for Machine Learning.

Development Tools

Amazon is a company that uses the learnings of open-source and AWS internal projects to give you tips on improving your code-base and infrastructure. They also provide all the tools for you to get started with software development.

Amazon CodeGuru is a Machine Learning-powered static code analysis tool currently available for Python and Java. It is trained on popular open-source repositories and internal Amazon repositories and gives you hints on improving your code based on industry best practices.

While CodeGuru does static analysis, Amazon DevOps Guru is a Machine Learning-enabled monitoring service for the cloud. It automatically monitors your infrastructure and provides alerts and insights for best practices.

The Text Tools

AWS has many Machine Learning-powered tools that help you understand, modify, and create texts. They are accommodating if you want to build conversational user interfaces or summarize texts.

Amazon Comprehend for natural language processing and text analytics helps you understand text sentiment and relate texts to each other.

Amazon Lex is a service for building conversational interfaces using voice and text. With Lex, you can use the same deep learning engine that powers Alexa in your own applications.

Amazon Textract extracts text and data from scanned documents. It's not just OCR but backed by Machine Learning models that have analyzed many types of documents, and can identify the contents of fields in forms and information stored in tables.

You can use Amazon Transcribe to turn any speech recording into a text. If you need to go the other way around, Amazon Polly will synthesize lifelike speech from any text.

Amazon Translate caters to your multilingual needs by translating every text into the language of your choice.

The Business Tools

The AWS Machine Learning tools with business focus help you with operational and sales tasks.

With the help of Amazon Forecast, the forecasting technology at the heart of Amazon.com, it is now possible to build forecasting models for your own applications.

Amazon Fraud Detector makes it easy to identify potentially fraudulent online activities such as online payment fraud and the creation of fake accounts.

Amazon Kendra is a Machine Learning-powered enterprise search service that helps your employees to find the data they need.

Amazon Personalize lets you personalize customer recommendations for your application with the same system used on Amazon.com.

Image & Video Tools

If you need to analyze visual data from videos or photographs, you will also find some Machine Learning support from AWS.

Amazon Rekognition can scan videos and images for objects and people. This allows you to verify, sort, and organize media based on their content alone.

AWS DeepLens is a hardware offering from AWS. It comes with a fully programmable camera you can use to train Machine Learning models for your specific task. Tutorials and guides also accompany this to get started right away.

AWS Panorama is a computer vision service that runs on the edge. This allows for use-cases that need low latency or don't have constant internet connections available.

Manufacturing Tools

For the future factory, AWS offers some entirely new Machine Learning solutions that can remove human error from manufacturing processes, be it finding defects in products or malfunctioning machinery.

Amazon Lookout for Vision spots product defects using computer vision to automate quality inspection.

Amazon Lookout for Equipment detects abnormal equipment behavior by analyzing sensor data.

Amazon Lookout for Metrics automatically detects anomalies in metrics and identifies their root cause.

Amazon Monitron is an end-to-end system that automatically detects abnormal behavior in industrial machinery, enabling you to take proactive action on potential failures and reduce unplanned downtime.

Other Tools

Besides the big Machine Learning tools we already listed, AWS also offers some more experimental audio and healthcare services.

Amazon HealthLake helps you to make sense of healthcare data. It's HIPAA-eligible, so your organization will be safe when handling sensitive data.

With AWS DeepComposer, you get an AI-enabled keyboard that transforms your melodies into unique songs with the help of Machine Learning models. While this goes astray of the well-known way artists create music, it can lead to new songwriting ways.

AWS DeepRacer gives you a 1/18th scale autonomous race car driven by Machine Learning models. You can train it at home, with your own parkour, and later compete in the global racing league to see if your training had any merit.

Summary

AWS has a huge catalog of Machine Learning services right at your fingertips with solutions for every stage of your process and different use-cases.

If you're already a Machine Learning professional, you can start with SageMaker to ease the burden of infrastructure provisioning in the cloud and scale-out to hundreds of machines. This way, you get quicker to the interesting part of your work and don't have to wait days to get things done.

If you just want to use the power of Machine Learning to get new insights on your already existing data without the need to build your own Machine Learning skills in-house, there are a multitude of solutions available for business, development, text, and image-based use-cases that could come up.

Further Reading:

Machine Learning with AWS Lambda

5 Must Have AWS Serverless Tools for your Starter Kit

Ultimate Guide to Monitoring Serverless Applications

Discussion (0)