At Intellify we’ve been proud to partner with AWS since we were founded in 2018.
Our data scientists have always enjoyed using AWS’ ML service – Amazon Sagemaker. It reduces the time they have to spend on admin and engineering by taking away many of the tasks required to build, train and deploy the custom machine learning models when working with our customers like Alinta, Henry Schein and Vodafone.
While our data scientists always enjoy building an end to end machine learning project in this way, there is still a reasonable amount of effort still required to get a useful model into production, so it normally requires a project running for a month or two. While some customers have unique problems which are a good fit for this kind of customized, flexible approach, we have seen some use cases arise regularly, particularly forecasting, personalisation and image processing.
We’ve been really excited to see AWS expand its AI services portfolio with Amazon Personalize and Amazon Forecast becoming generally available this year. It means that for the most common ML use cases we can often dramatically reduce the work to do in areas like:
Exploratory data analysis
Model training and deployment
ML pipeline engineering
Production system monitoring
This means we can deliver projects at a fraction of the cost (both professional services and AWS platform) and time which was previously required. If the data is already on AWS, it’s often possible now to test out ML in hours to validate whether it’s a good use case.
At the moment we’re using Amazon’s AI Services including Amazon Forecast, Amazon Personalize, Amazon Rekognition and Amazon Comprehend Medical in use cases like
Inventory planning and control
Business metric forecasting
Removal of PII from images and reports
It’s amazing how many features these products have and how quickly we can get the insights we need to impact key business processes.
We’re enjoying using these services and have some packaged projects defined to quickly deliver outcomes in the use cases above based on our experience working with AWS and their AI Services. If you’re interested in proving out ML use cases rapidly and at low cost, we’d love to talk.