Topic 1, Trey Research
Background
Trey Research is a technology innovator. The company partners with regional transportation department office to build solutions that improve traffic flow and safety.
The company is developing the following solutions:

Regional transportation departments installed traffic sensor systems on major highways across North America. Sensors record the following information each time a vehicle passes in front of a sensor:
Time
Location in latitude and longitude
Speed in kilometers per second (kmps)
License plate number
Length of vehicle in meters
Sensors provide data by using the following structure:

Traffic sensors will occasionally capture an image of a vehicle for debugging purposes.
You must optimize performance of saving/storing vehicle images.
Traffic sensor data
Sensors must have permission only to add items to the SensorData collection.
Traffic data insertion rate must be maximized.
Once every three months all traffic sensor data must be analyzed to look for data patterns that indicate sensor malfunctions.
Sensor data must be stored in a Cosmos DB named treydata in a collection named SensorData The impact of vehicle images on sensor data throughout must be minimized.
Backtrack
This solution reports on all data related to a specific vehicle license plate. The report must use data from the SensorData collection. Users must be able to filter vehicle data in the following ways:
vehicles on a specific road
vehicles driving above the speed limit
Planning Assistance
Data used for Planning Assistance must be stored in a sharded Azure SQL Database.
Data from the Sensor Data collection will automatically be loaded into the Planning Assistance database once a week by using Azure Data Factory. You must be able to manually trigger the data load process.
Privacy and security policy
Azure Active Directory must be used for all services where it is available.
For privacy reasons, license plate number information must not be accessible in Planning Assistance.
Unauthorized usage of the Planning Assistance data must be detected as quickly as possible. Unauthorized usage is determined by looking for an unusual pattern of usage.
Data must only be stored for seven years.
Performance and availability
The report for Backtrack must execute as quickly as possible.
The SLA for Planning Assistance is 70 percent, and multiday outages are permitted.
All data must be replicated to multiple geographic regions to prevent data loss.
You must maximize the performance of the Real Time Response system.
Financial Requirements
Azure resource costs must be minimized where possible.