Low Latency Live Streaming Pipeline Deployment Spec
Architect a zero-downtime deployment specification for live OTT video ingest, transcoding, and multi-CDN distribution.
Use this template when designing or upgrading continuous delivery workflows for broadcast-grade live video streaming platforms. It delivers an actionable engineering specification covering automated canary rollouts, stream-health telemetry gates, and edge cache invalidation.
Role: Principal Video Infrastructure Engineer with 12+ years designing broadcast OTT pipelines and zero-downtime CI/CD workflows.
Context
- Platform Name: {{streaming_platform_name}}
- Latency Budget: {{target_latency_budget_ms}}
- CI/CD Toolchain: {{current_ci_tooling}}
- Multi-CDN Edge Network: {{cdn_providers_list}}
- Video Transcoding Engine: {{transcoding_engine}}
- Regulatory Compliance: {{compliance_standard}}
Task
Generate a comprehensive technical deployment specification for {{streaming_platform_name}} that automates pipeline updates to the {{transcoding_engine}} ingest and packaging layer without dropping active live broadcast frames or exceeding {{target_latency_budget_ms}}.
Method
- Analyze {{transcoding_engine}} deployment requirements across stateful ingest nodes and stateless packagers.
- Formulate blue-green rollout stages inside {{current_ci_tooling}} to switch active manifest generation safely.
- Establish edge pre-warming and purge sequencing across {{cdn_providers_list}} to eliminate cache poisoning during container cutover.
- Define synthetic stream probe verification gates measuring end-to-end latency against {{target_latency_budget_ms}}.
- Design automated rollback triggers driven by RTP/HLS segment drops, frame synchronization drift, and 5xx response thresholds.
- Detail secrets injection and encryption key handling compliant with {{compliance_standard}}.
- Produce infrastructure-as-code parameter maps and pipeline execution state diagrams.
Constraints
- MUST guarantee continuous stream availability with zero dropped video chunks during deployment.
- MUST NOT require manual DNS or CDN switching during standard automated releases.
- All telemetry checks must complete within 90 seconds per canary stage.
- Pipeline security must satisfy {{compliance_standard}} credential segregation policies.
Output format
Provide a technical specification structured under the following exact markdown headers:
- Pipeline Architecture & Canary Topology (max 300 words)
- Automated Stage Matrix & Health Gates (table with Stage, Action, Gate Condition, Failure Action)
- Edge Caching & Multi-CDN Synchronization Spec (max 250 words)
- Automated Rollback & Disaster Recovery Protocol (ordered procedure)
Self-review
- Confirm all 6 context variables are explicitly addressed in the deployment strategy.
- Verify that latency constraints strictly adhere to {{target_latency_budget_ms}}.
- Validate that no manual intervention steps exist in the primary rollout path.
Explicit role, a named task, and discrete steps the model can follow.
Background, inputs and variables the model needs before it starts.
Hard boundaries — what the model must and must not do.
A named, field-level shape for the response.
Ordered work items that force analysis before an answer.
Length and structure that travel across frontier models.
Signal density — instruction weight without padding.
Documented variables so the scaffold adapts to new inputs.
Quality bar, assumptions and behaviour when inputs are thin.
How much real usage the template has behind it.