The Discovery-to-Iteration Lifecycle Loop
CoreSupport the solution across the discovery-to-iteration lifecycle · Difficulty 3/5
0%
lifecyclemonitoringiterationhandofffeedback-loop
Prerequisites
Explanation
An architect supports the solution across its whole life, not just at design time. The lifecycle has five named phases:
| Phase | Architect's role |
|---|---|
| Discovery | Elicit and validate requirements and constraints |
| Design | Choose patterns, models, integration, and governance; document tradeoffs |
| Handoff | Transfer the design with documentation and implementation guidance |
| Monitoring | Ensure observability/eval are in place; watch quality, latency, and cost against SLAs |
| Iteration | Feed monitoring and stakeholder feedback back into improvements |
A Loop, Not a Line
The critical mental model is that the lifecycle loops: monitoring and iteration feed back into design as requirements, data, and models evolve. It is not a one-directional pipeline that terminates at handoff. This mirrors the feedback-loop discipline of stakeholder communication and the monitoring practices used elsewhere in production architecture -- the architect keeps the loop turning across all of them.
Why Each Phase Depends on the Ones Before It
- Design depends on discovery's agreed outcome and constraints
- Handoff depends on design's documentation and rationale
- Monitoring depends on handoff having included observability/eval setup
- Iteration depends on monitoring surfacing real signal, and on the stakeholder feedback loop
Common exam traps
- Treating design as "done" at handoff. Non-determinism, model updates, and data drift mean the architecture needs ongoing monitoring and iteration to stay within its SLAs. When monitoring reveals latency creeping past an agreed SLA after a data-volume increase, the correct move is to feed that finding back into iteration and re-align expectations with stakeholders -- not to do nothing because the design was signed off at handoff, disable monitoring to silence the alerts, or reach for an unrelated parameter tweak.
Key Takeaways
- The five lifecycle phases are discovery, design, handoff, monitoring, iteration
- The lifecycle is a loop: monitoring and iteration feed back into design as requirements, data, and models evolve
- Handoff requires documentation plus implementation guidance; monitoring requires observability/eval to already be in place from handoff
- Classic trap: treating design as "done" at handoff -- non-determinism, model updates, and data drift demand ongoing monitoring and iteration
- When monitoring shows an SLA breach, feed it into iteration and re-align with stakeholders, don't ignore it or silence the alert
Glossary Terms
Related Concepts