Serverless deployment

Intent

Deploy services as managed functions where the platform handles scaling and runtime management.

Use when

• Workloads are bursty and fit within execution limits; you want reduced operational burden.
• You can tolerate platform constraints (cold starts, limited runtime control).

Avoid / watch-outs

• Cold starts and concurrency limits can impact latency; be explicit about SLOs.
• Observability and local debugging can be harder; invest in tracing/logging.

Skill mapping

• spec: define latency and availability requirements under cold starts/bursts.
• observability: ensure traces/logs/metrics work in the serverless environment.
• resilience: time budgets and retry strategy aligned to function execution limits.