Domain 2.2: Design data integration

A data engineering team is building a modern data warehouse in Azure. They have 500 TB of historical telemetry data stored in Parquet format in Azure Data Lake Storage Gen2. Data analysts need to run ad-hoc SQL queries against this data using their existing BI tools (like Power BI and Tableau). The queries are run infrequently (a few times a week), but when they run, they require massive compute power. The company wants to pay only for the queries executed, rather than provisioning dedicated compute clusters 24/7. Which Azure Synapse Analytics component should you recommend?

Your company is designing an ELT (Extract, Load, Transform) pipeline. The pipeline must extract data from 50 different on-premises and cloud data sources, load the raw data into Azure Data Lake Storage Gen2, and then perform complex transformations using Apache Spark. The data engineering team prefers writing transformation logic in Python and Scala notebooks. Which TWO Azure services should you combine to build this solution? (Select TWO)

An IoT architecture receives millions of temperature readings per second from sensors worldwide. You need to design a solution to ingest this massive stream of data and immediately calculate the average temperature over a 5-minute tumbling window. If the average exceeds a threshold, an alert must be sent to a Service Bus queue. The solution must be fully managed and require minimal custom code. Which combination of services should you recommend?

Your company receives 500 GB of raw JSON telemetry data daily from IoT devices, stored in Azure Data Lake Storage Gen2. You need to design a data integration solution to transform this data into a structured format for reporting. The transformation logic requires complex data cleansing using Python and Spark. The solution must support orchestration of the pipeline and minimize cluster startup times. Which Azure service should you recommend?

You are designing a real-time fraud detection system for a financial services company. Millions of credit card transactions per minute will be ingested. The system must evaluate these transactions against a machine learning model in real-time. Transactions flagged as fraudulent must be sent to a Service Bus queue for immediate human review. All raw transactions must be archived to Azure Data Lake Storage Gen2 for future model training. Which TWO Azure services should you include in the data integration architecture? (Select TWO)

Your data engineering team is building a 'Lakehouse' architecture using Azure Databricks. They need a storage format that supports ACID transactions, scalable metadata handling, and unifies streaming and batch data processing. It must also allow them to perform 'time travel' queries to view previous versions of the data. Which data format should you recommend?