Advanced College of Engineering - AWS AI Day

AWS AI Day was a single session delivered on campus at Advanced College of Engineering and Management, on the artificial intelligence and machine learning services AWS provides and what can be built with them.

The AWS AI catalogue sits at three levels, and knowing which level a problem belongs to is most of the decision. At the top are the ready-made services — Rekognition for images and video, Transcribe and Polly for speech, Comprehend for language, Textract for documents — which solve a defined problem behind an API call and need no training data. Below those sits Amazon Bedrock, which gives access to foundation models without managing infrastructure. Below that again is SageMaker, where a model is trained, tuned and deployed when nothing off the shelf will do.

The gap this addresses is a familiar one. Students often meet machine learning as mathematics and then never see how it is packaged, paid for and operated in production — which problems are worth a managed service, what a foundation model changes about the build-versus-buy question, and how much of the real engineering work sits around the model rather than inside it.