System Design: Company-Wide Scheduler Platform and Employee Review & Rewards System
You are designing two platforms for a fast-growing, multi-tenant technology company. Address each part independently, but note any shared building blocks if relevant.
Part A — Company-Wide Scheduled Job Platform (Cron-like)
Design a horizontally scalable platform to run scheduled workflows across services.
Include:
-
Scheduling and execution
-
Schedule management: crontab expressions, calendar/holiday exceptions, and time zones (DST-safe)
-
Workflow DAGs and task dependencies
-
Retries with exponential backoff and idempotency guarantees
-
Backfills for missed/retroactive runs
-
Concurrency control: global, per-tenant, per-workflow limits
-
Access control and multi-tenant isolation
-
Reliability and operations
-
Observability: logs, metrics, traces; failure alerting
-
Horizontal scalability: sharding and stateless workers
-
Storage schemas for schedules, DAGs, runs, leases/locks, and audit
-
Coordinator/worker architecture: leader election, queues, heartbeats
-
Disaster recovery and operational runbooks
Discuss trade-offs (e.g., at-least-once vs exactly-once), and provide concrete storage schemas and component interactions.
Part B — Employee Review & Rewards System
Design a system that supports performance reviews and employee rewards with strong privacy and compliance.
Include:
-
Core features
-
Performance review cycles (self/manager/peer), calibration, and peer feedback
-
Reward points/grants, redemption, policies, and approval workflows
-
Privacy/visibility controls and audit trails
-
Architecture and integrations
-
Service boundaries and data models
-
Integrations with HRIS (org, titles, managers), SSO, and payroll/stock systems
-
Abuse/fraud prevention, rate-limiting, and anomaly detection
-
Operations
-
Observability, SLAs, and disaster recovery
-
Data retention, legal holds, and regional data residency
Provide data schemas, state machines, and key workflows. Call out edge cases, safeguards, and scale assumptions.