Skip to content
AAOE
Draft2 min read

Discovery Lifecycle

Purpose

Engineering discoveries exist so that observations made during real work can become reusable organizational knowledge rather than being lost once the task that produced them is done. A Discovery is the first formal representation of newly acquired engineering knowledge: it precedes architectural decisions, specifications, and implementation.

A Discovery is not yet architecture and not yet a standard. It is an observation, captured and preserved so it can later be evaluated, validated, and — if warranted — formalized.

Principles

  • Discoveries originate from evidence, not speculation.
  • Discoveries describe observations before prescribing solutions.
  • Multiple discoveries may lead to one architectural decision.
  • Discoveries remain traceable to the work that produced them.
  • Discoveries preserve engineering learning, whether or not they are ultimately formalized.

Discovery Sources

Discoveries may originate from:

  • Practical implementation
  • Experiments
  • Production incidents
  • Architecture reviews
  • User feedback
  • Performance investigations
  • Research

Discovery Lifecycle

Observation

Discovery

Validation

Architectural Decision

Specification
  • Observation — something notable is encountered during engineering work, from any of the sources above.
  • Discovery — the observation is captured as a Discovery: recorded, evidence-based, and traceable to its source.
  • Validation — the Discovery is examined for reproducibility and general applicability beyond the situation that produced it.
  • Architectural Decision — a validated Discovery may lead to an architectural decision that evaluates and responds to it.
  • Specification — an accepted architectural decision may in turn be formalized into a specification.

Discovery Criteria

A Discovery should:

  • Be evidence-based.
  • Be reproducible.
  • Be generally applicable.
  • Be independent of tooling whenever possible.

Outputs

  • Candidate architectural decisions.
  • Candidate specifications.
  • Engineering insights.
  • Organizational learning.

Governance

  • Discoveries are not automatically architecture.
  • Discoveries do not change specifications directly.
  • Architectural decisions evaluate discoveries; a Discovery does not carry authority on its own.
  • Specifications formalize approved discoveries, not raw observations.

Future Evolution

Discovery management may evolve — for example, in how discoveries are recorded, validated, or connected to downstream decisions — while preserving evidence-first engineering as its foundation.