System Architecture & Design Flashcards
7 cards from real TOC practice questions. Tap to flip, then mark Knew It or Still Learning — missed cards come back until you master them.
Read the first 7 System Architecture & Design flashcards as text
A TOC is designing a predictive hiring model. Which validation technique ensures the model performs well on unseen candidate data?
Answer: Cross-validation with held-out test sets
Cross-validation tests model performance on data not used in training, revealing how well it generalizes to new candidates.
In talent system architecture, 'eventual consistency' means:
Answer: Data replicas will converge to the same state given sufficient time
Eventual consistency allows distributed systems to accept writes immediately while guaranteeing replicas will converge to the same value over time.
Which system architecture component is responsible for enforcing data validation rules when talent assessment responses are submitted?
Answer: Business logic / application layer
The business logic layer validates input data against defined rules before it is persisted to the database.
A talent platform needs to support 10,000 concurrent assessment sessions during peak hiring season. Which architecture pattern best addresses this scalability need?
Answer: Horizontal scaling with load balancing
Horizontal scaling adds server instances behind a load balancer, distributing concurrent users across the cluster without a single bottleneck.
When integrating talent behavioral data with performance management systems, data normalization ensures:
Answer: Consistent data formats and scales across source systems
Data normalization standardizes formats, units, and scales so data from different systems can be accurately compared and combined.
A talent optimization system uses a 'strangler fig' migration pattern. This describes:
Answer: Incrementally replacing legacy components with new ones over time
The strangler fig pattern gradually replaces legacy functionality with new implementations, reducing migration risk by transitioning piece by piece.
In a talent analytics platform, a 'data warehouse' differs from an 'operational database' primarily in that it is:
Answer: Optimized for large-scale historical queries and reporting
Data warehouses are designed for analytical queries across large historical datasets, whereas operational databases handle real-time transactional workloads.