I was excited to join a group of Toads attending AWS re:Invent.
How to do machine learning in AWS
After being immersed in hundreds of announcements and innovations and inspiring presentations at AWS re:Invent, the key insight I came away with was this: machine learning has so radically evolved
After being immersed in hundreds of announcements and innovations and inspiring presentations at AWS re:Invent, the key insight I came away with was this: machine learning has so radically evolved that it will change the future of every industry smart enough take advantage of it.
There are tons of great new services and features, but what’s really exciting to me—and promising for Metal Toad’s clients—is the big picture of AWS Machine Learning.
It’s been around for a while, but AWS continues to add services and features that have made it a genuine game-changer. It’s like a data scientist in a box.
Where once you had to invest in scientists and programmers to code from scratch—not to mention the servers and infrastructure costs—now you can work with an AWS Consulting Partner who does the massive legwork of understanding and applying the tech, leverage AWS’s automated machine learning, and pay for serverless infrastructure by the second. Just implement SageMaker, and you’ve got a wildly powerful machine learning system at your fingertips. Not only is the tech better, it’s cheaper—so much so that large enterprises will save millions of dollars.
Enterprises have a massive amount of data, and once you have a product like SageMaker, the possibilities for leveraging that data are nearly endless. My fellow Toads have talked about the huge potential of some SageMaker microservices—Forecast, Personalize, Rekognition—and I’m just as enthusiastic as they are. But that’s just the tip of the iceberg. Fraud detection, smart search, AI chatbots, transcription, user behavior analysis, and more are all powerful aspects of AWS’s machine learning suite—there’s even machine learning tech that will automatically decide which machine learning service you need!
Just one example of a huge recent advancement. This is truly machine learning on a grandiose scale. Comprehend analyzes deduces human expression in text in a totally unique way. For instance, you could feed it the comments on your org’s Facebook page and run a sentiment analysis to understand instantly how people feel about your brand—the ultimate data-validated gut check.
With such rapidly evolving tech and endless possibilities, it’s easy to go down the rabbit hole of AWS machine learning services—I often find myself there, as a matter of fact! But I always go back to the big picture—what’s it all mean for the clients and industries we support? We’re witnessing a revolution in automation, but really harnessing the power of these advancements requires us to expand our idea of what automation really means. One of my somewhat controversial points of view is that most marketing automation tools in use today are ultimately a waste of money. Why? They’re automating actions—sending emails, pushing ads, filling social feeds—instead of automating insights.
The mass of data available to your enterprise is worthless until it’s analyzed and translated into actual useful information which can be kicked off by finding the right AWS managed services vendor. Once on the path, machine learning becomes truly revolutionary when it turns data into meaning—business intelligence you can use to make better decisions and optimize your business performance.
Artificial intelligence is machines doing stuff—augmented intelligence is machines turning data into the kind of insight humans can use to make faster, better, smarter decisions.
Because the future isn’t about the machines. It’s about how we use them to create new human realities.