2 3 4 5 Build a low-latency data pipeline from multiple, geographically dispersed on-premises or cloud-based databases using AWS global infrastructure and fully managed networking and data synchronization services. Alternatively, migrate existing proprietary database to the AWS Cloud to reduce cost, increase security and performances. Stream and transform live data in real time with Amazon Kinesis. Store all the data in a single data lake on Amazon S3, achieving improved durability, redundancy and security. Prepare data for machine learning (ML) processing with AWS Glue and store Amazon SageMaker ML models and outputs into dedicated S3 buckets. Use S3 intelligent tiering and storage classes to achieve scalability at a fraction of the cost of an on-premises solution. Orchestrate Amazon SageMaker capabilities with a manual or an automated workflow. Train, evaluate, Optimize, deploy, and test the ML model. Use SageMaker to expose your model as an API and run real-time or batch predictions. Use a largely serverless architecture to be more cost efficient. Enable real-time push notifications to your customers and partners when SageMarker detects possible vessel delays. Use an AWS Lambda function and SageMaker endpoint integration with Amazon Kinesis to raise an alarm for the operator each time a delay situation is detected. Import the predictions to enrich your Amazon Redshift data warehouse. Visualize the reports using Amazon QuickSight or other reporting tool to discover trends and implement corrective actions
