A company's near-real-time streaming application on AWS ingests data and runs a job that takes 30 minutes to complete. The workload often suffers high latency from large incoming data volumes. A solutions architect needs a scalable, serverless design to improve performance. Which combination of steps should the architect take? (Choose two.)

Amazon Kinesis Data Firehose is the correct choice for data ingestion because it is a fully managed, serverless service designed for near-real-time streaming data delivery to various AWS destinations. It automatically scales to handle large data volumes, addressing the high latency issue. AWS DMS is incorrect as it's for database migration, not general streaming data ingestion. For processing, AWS Fargate with Amazon ECS is correct. It provides a serverless compute engine for containers, allowing the application to scale automatically without managing servers, ideal for a job that takes 30 minutes and needs to handle varying loads. AWS Lambda with Step Functions is less suitable for a 30-minute job, as Lambda has a maximum execution duration of 15 minutes. EC2 instances in an Auto Scaling group are not serverless, requiring server management.