A framework for preserving intent

Iterative systems
naturally lose intent.
IDD keeps it alive.

Every handoff, iteration, and optimization cycle introduces interpretation. Requirements evolve. Code evolves. Designs evolve. But the original why gets buried. Intent Driven Design treats intent as a first-class artifact throughout the entire delivery lifecycle.

01
Original Intent
The purpose. The why.
02
Iteration
Cycles of delivery
03
Interpretation
At every handoff
04
Optimization
For local goals
05
Intent Drift
The gap that grows
IDD
Clarity Point
Intent restored
IDD animated explainer: from original intent through iterations, drift, clarity point, to intent fidelity
The Original Insight

The problem isn't Agile.
It isn't AI. It's every iterative system.

IDD started with a simple observation: every time information passes from one person, team, artifact, or iteration to another, some of the original intent is lost.

In SAFe, this happens at every layer — from Executive Vision down through Portfolio, Epic, Feature, Story, Code, and Release. At each handoff, people interpret intent differently, teams optimize for local objectives, assumptions are introduced, and context disappears.

By the time a solution reaches production, the organization may have delivered exactly what was requested in the backlog — while failing to deliver what was originally intended.

“A project can have green status, healthy velocity, and successful releases — while simultaneously experiencing significant intent drift.”

SAFe was simply the first environment where this was observed. The same pattern exists in AI-assisted development, enterprise transformation, product development, and agentic systems. The root cause is always the same.

IDD's answer: treat intent as a first-class artifact — not something buried in a kickoff deck, but something explicitly named, carried forward, and continuously validated throughout the lifecycle.

Layer
What gets lost
Risk
Executive Vision
Strategic context filtered by portfolio priorities
Low
Portfolio → Epic
Mission rationale compressed into business outcomes
Med
Epic → Feature
Operational purpose abstracted into delivery scope
Med
Feature → Story
User value decoupled from originating objective
High
Story → Code
Intent replaced by implementation assumptions
High
Code → Release
Original purpose rarely validated at deployment
High
The Framework

Four constructs. One goal:
keep intent alive.

IDD is built on four distinct constructs — a principle, a mechanism, a measure, and a practice. Together they form a lightweight intent-preservation layer that sits on top of whatever delivery framework you already use.

Principle
Mechanism
Measure
Practice
01 — Principle

Intent as a first-class artifact

Original intent must be captured, named, and carried forward — not left in a kickoff meeting or business case. It is a living artifact, not a historical document.

02 — Mechanism

Clarity Points

Deliberate moments throughout the lifecycle where teams resurface the original intent and ask: does what we're building still reflect why we started?

03 — Measure

Intent Fidelity

Intent Fidelity is the degree to which a solution remains aligned with its intended purpose as it evolves. It is measurable, trackable, and independent of velocity or quality metrics.

04 — Practice

Lean Assessments

Lightweight alignment checks — not reviews or approvals — that evaluate whether execution has drifted from the original operational or business objective.

Intent Fidelity

Every iteration has a position.
Fidelity is the distance from intent.

Fidelity is not about activity. It's about proximity to purpose. The closer your solution stays to the original intent, the higher the fidelity — and the more likely you are to deliver real value. Clarity Points exist to re-center the work and restore fidelity when drift occurs.

The Intent Fidelity Map: concentric rings showing how iterations drift from original intent, with Clarity Points re-centering toward high fidelity. Includes the Fidelity Zone legend (High, Medium, Low, Very Low) and a four-step diagram of how a Clarity Point restores intent alignment.
Click to view fullscreen
Applications

One core idea.
Two proven applications.

IDD emerged from SAFe enterprise delivery. AI-assisted development brought it home — because AI amplifies the same drift pattern at an order of magnitude greater speed.

SAFe & Agile Delivery

Large-scale programs create many abstraction layers. Each layer introduces interpretation. IDD reconnects every work product back to the originating mission intent.

Product VisionOperational Intent
PI ObjectivesOutcome Alignment
UATOperational Validation
Iteration ReviewsClarity Evaluation
I&A EventsLean Assessment
Release GovernanceAlignment Visibility

AI-Assisted Development

AI accelerates iteration to the point where intent degradation can occur within hours. IDD operates as an intent-preservation layer across prompts, generated code, refactoring, and agent decisions.

Prompt evolutionIntent anchor validation
Generated codeFidelity check
Auto-refactoringObjective alignment review
Agent decisionsClarity Point checkpoint
Optimization cyclesMission-outcome test
Agentic pipelinesHuman intent preservation
Clarity Points

Not checkpoints.
Not reviews. Intent resurface moments.

Clarity Points are not compliance gates. They are deliberate pauses where teams ask the questions that delivery frameworks never ask.

“What was the original objective?”

Not the current ticket. Not the sprint goal. The original reason this work was initiated — what problem it was created to solve.

“What assumptions have changed?”

Assumptions made at the start are rarely revisited. A Clarity Point forces teams to surface what has shifted since intent was first defined.

“Does the current solution still reflect the original purpose?”

Technically correct, on-time, on-budget — but are we still building the right thing? This is the question most delivery frameworks never ask.

“Have we optimized ourselves away from the original need?”

Performance, cost, and efficiency improvements can silently shift a solution away from its founding purpose. Optimization is productive drift.

Intent fidelity across iterations — Clarity Points restore alignment
Start
100%
Iter 1
88%
Iter 2
74%
Iter 3
61%
CP ①
94%
Iter 5
83%
Iter 6
69%
CP ②
91%
High fidelity
Drifting
Clarity Point — intent restored
Community

The theory is evolving.
Your experience shapes it.

IDD is a living framework, not a finished one. Whether you're navigating SAFe program complexity or AI-assisted delivery, your real-world observations help define what comes next.

Share a use case

Where have you seen intent drift in your own programs? What triggered it? How did you catch it?

Challenge the framework

Where does IDD break down? What edge cases hasn't it addressed? Push back — the theory gets stronger for it.

Suggest an application

Beyond SAFe and AI — where else does iterative drift occur? Product development? Policy? Research? Tell us.

“Is the original intent still present?”

That is the question traditional delivery frameworks were never designed to answer. IDD exists to answer it — continuously, lightly, and without replacing the systems you already depend on.

Read the Theory