documentation/docs/strategy/target-operating-model.mdx

---
id: target-operating-model
title: Target operating model
sidebar_label: Target
description: An autonomous agent executes and governs the SDLC; humans direct, guide and review the repeating cycle.
hide_table_of_contents: true
---

import ArchifyDiagram from '@site/src/components/ArchifyDiagram';

<div className="strategy-content">

**TO-BE · Target operating model — the intended strategy, not a deployed system.**

## Autonomous SDLC, human direction

<ArchifyDiagram name="strategy-target" title="Autonomous agent executes and governs the SDLC; humans direct, guide and review" story />

An autonomous agent executes and governs every phase. Humans direct, guide and review it; a separate human-controlled AI Agent contributes to each phase from outside the cycle.

**Solid arrows:** lifecycle · **Dotted spokes:** execution and governance · **Dashed arrows:** contributions.

**Scope → Architecture → UI/UX → Planning → Delivery → UAT → Security → Releases → Scope**

Each stage follows the previous one. Releases returns to Scope; the cycle repeats indefinitely, with human direction informing the next iteration.

<details>
<summary>Supporting implementation options: trust boundaries and recovery</summary>

## Controls that support autonomous execution

The lifecycle above is the strategic direction. The earlier control design below is a supporting implementation option, not an agreed assignment of approval roles or a claim of installed controls. Detailed approval policy and responsibilities remain to be defined with humans.

1. Humans direct, guide and review; the autonomous agent executes and governs work across the lifecycle.
2. The reviewed registry supplies pinned skills and qualified adapters to coordinator and local clients.
3. Laptop contributors can participate alongside the agent; independent repository and CI checks verify the exact revision.
4. A separate release service and artifact-scoped approval are possible controls, subject to the policy humans define.
5. Restricted audit and measurement evidence supports human review and the next Scope iteration.

<ArchifyDiagram name="ai-adoption-target" title="Supporting control option, not agreed approval policy: isolated release authority" />

**TO-BE · Target operating model.** Introduction depends on the [roadmap gates](./roadmap.mdx); these pages do not describe installed controls or measured company results.

### One coordination function, distributed contributions

The autonomous agent is more than a task router: it executes and governs the repeating SDLC under human direction, guidance and review. A coordinating function maintains work records, dependencies, evidence requests, bounded retries and escalation. Data & AI and the existing AI Architect function are possible implementation contributors, not newly assigned approval authorities. Autonomous execution does not remove human involvement or allow the agent to redefine its own controls.

One logical coordinator does not mean one fragile process. A durable task ledger, explicit state transitions, leases and idempotency keys allow recovery or multiple replicas without duplicate delivery. Inference outages stop new agent decisions, not the organization's ability to take over. People can resume work through the normal deterministic workflow. Evidence and authorization services fail closed for privileged actions; missing evidence never becomes permission.

These controls support the sequence Scope → Architecture → UI/UX → Planning → Delivery → UAT → Security → Releases → Scope; they do not replace it with a separate lifecycle. Product, engineering, design, quality, security and platform expertise inform human direction and review. Specific approval assignments and release permissions are implementation decisions still to be defined.

### Trust and authority boundaries

People authenticate to the work system; the coordinator receives only the project context it needs. Untrusted issue text, repository content, web pages and tool output are task data, never an authorization source. Local clients operate under device controls and project-scoped access. Hard containment requires OS/container isolation, filesystem and network restrictions and separate short-lived identities; a skill folder, prompt or working directory is not a sandbox.

The registry distributes reviewed immutable content and adapters, not ambient credentials. Executable scripts, plugins, MCP servers and dependencies are separately reviewed and pinned. Model-provider routing is allowlisted for the data classification, with approved retention and training-use terms before sensitive data crosses that boundary. Fallback to another provider must not silently change the data policy.

Central controls verify protected-branch review, actual artifact identity, exact commit and policy revision, required checks and a scoped human approval. Candidate changes cannot edit their own authoritative gate or reuse approval for a different artifact. Deployment credentials belong to the release service, not the local agent. Approval expires or is invalidated by changed content. Emergency stop revokes mutation credentials and stops scheduling independently of the model, preserves evidence and enables human takeover.

</details>

<details>
<summary>Candidate contributions and first workflows for all eleven teams</summary>

### Department contributions

The existing team structure provides a starting point. The contributions below are implementation proposals, not settled approval roles or lifecycle gates.

### Product / Product Management

**Contribution and retained human decision** — Own pilot priorities, acceptance criteria and outcome hypotheses; decide product value.

**First bounded workflow and evidence** — Draft acceptance criteria from sanitized discovery notes; independent completeness review and accepted-task success.

### Product / Product Design

**Contribution and retained human decision** — Own design quality, accessibility and research handling.

**First bounded workflow and evidence** — Review a proposed flow against an accessibility checklist; verify findings against the interface.

### Technology / Product Engineering

**Contribution and retained human decision** — Own skill authoring, code review and maintainability; remain accountable for generated code.

**First bounded workflow and evidence** — Test scaffolding and change summaries in a disposable branch; independent tests and escaped-quality follow-up.

### Technology / Platform & IT

**Contribution and retained human decision** — Own supported client inventory, distribution, identity, isolation, CI/release integration, transport and emergency revocation.

**First bounded workflow and evidence** — Qualify a pinned bundle and exercise rollback; inventory and exact-revision gate evidence.

### Technology / Architecture & Security

**Contribution and retained human decision** — Own risk classification, data boundaries, security review and bounded exceptions.

**First bounded workflow and evidence** — Review injected-input and unauthorized-tool cases; independent control evidence and policy-effect counts.

### Technology / Quality Engineering

**Contribution and retained human decision** — Own regression/adversarial suites, rubric calibration and acceptance evidence.

**First bounded workflow and evidence** — Evaluate candidate versus approved skills on repeated representative tasks; retain all failures.

### Technology / Data & AI

**Contribution and retained human decision** — Own catalog coordination, coordinator evaluation and measurement design, through existing AI enablement responsibilities.

**First bounded workflow and evidence** — Curate a small skill set and analyze a prospective pilot; no new AI department.

### Operations / Operations Squad

**Contribution and retained human decision** — Own service acceptance, disputes, fraud, refunds and customer communications.

**First bounded workflow and evidence** — Draft an answer from sanitized resolved cases; Operational Quality workstream checks correctness before any send. No automated dispute/refund decision.

### Marketing / Marketing Squad

**Contribution and retained human decision** — Own brand, factual claims, content rights, publication and spend approvals.

**First bounded workflow and evidence** — Draft a campaign brief against approved brand material; human claim/rights review before publication.

### Finance & Accounting / Finance & Accounting Squad

**Contribution and retained human decision** — Own allocation of total costs, financial evidence and segregation of approvals.

**First bounded workflow and evidence** — Explain a sanitized reconciliation exception; no posting, payout or payment authority.

### HR & Administration / People & Administration Squad

**Contribution and retained human decision** — Own coaching, acceptable use, feedback privacy and people decisions; partner with Platform & IT on access lifecycle.

**First bounded workflow and evidence** — Draft onboarding guidance from approved policy; human accuracy review. No hiring or performance decision automation.

Product keeps two teams; Technology keeps five; the other four departments keep one squad each. Coordination, skill ownership, approval and cover are assignments inside these teams, not new standing departments, roles or headcount. Department leaders retain accountability and designate real approvers before a pilot starts.

See the existing [organization structure](/organization/structure/), [department collaboration](/organization/collaboration/) and [SDLC responsibilities](/organization/sdlc/).

</details>

<details>
<summary>Strategy rationale and reading map</summary>

### Govern skills first, expand authority only with evidence

Eldorado's target is to make AI-assisted work dependable across the organization: useful guidance, clear human ownership, controlled execution and measurable outcomes. The first investment is a governed skills system, not an autonomous replacement for the organization. A skill packages reusable working knowledge, prerequisites and verification steps. It helps an agent perform a task; it does not grant permission or guarantee that the work is correct.

The target combines an autonomous agent executing and governing the SDLC with humans directing, guiding and reviewing it. Developer choice on local laptops remains a supporting option: Copilot, Claude Code and Cline can receive reviewed portable skill content through qualified adapters. Repository, CI and release controls support execution; detailed approval policies remain implementation decisions, not new roles established by this strategy.

The strategy describes the intended operating model and the evidence required to introduce it. It does not establish that these controls are installed, that every client is supported, or that AI has improved Eldorado's performance. The current reference is the documented six-department organization. Rollout begins by validating actual clients, responsibilities, data boundaries and available evidence.

Success means accepted outcomes with maintained quality, controlled risk and understood total cost. Faster activity without reliable outcomes is not success. More skill invocations, generated lines or tool subscriptions are not productivity targets. People remain accountable for product, customer, financial and personnel decisions.

### Navigate the strategy

- [Roadmap](./roadmap.mdx) — Evidence-gated phases, no promised dates or budgets.
- [Skills governance](./skills-governance.mdx) — Contract, lifecycle, decision rights, adapters, locking, exceptions.
- [Skills quality](./skills-quality.mdx) — Risk tiers, independent evaluation, repeated trials, adversarial conformance.
- [Skills adoption](./skills-adoption.mdx) — Workflow-led coaching, client choice, feedback, manual path.
- [Measurement](./measurement/index.mdx) — Full KPI contracts, minimal events, illustrative calculations, privacy and limits.
- [Skills maintenance](./skills-maintenance.mdx) — Review, drift, deprecation, rollback and incident handling.

</details>

**Next:** [Follow the Roadmap toward Target →](./roadmap.mdx)

</div>