How we do it

From AI ambition to working outcomes.

Start with the decision in front of you—not a framework name. Choose one of seven situations to see what you may already have, what needs deciding and which existing method can help. Strategy, operations and implementation are context, not a required sequence.

Find your situation
  1. When the opportunity is still taking shape We find the value worth pursuing.
  2. When good ideas pull in different directions We help you choose the right bets.
  3. When the strategy is already there We make your strategy work.
  4. When a process needs a better way Together, we improve the work.
  5. When a pilot is ready for its next decision We help pilots earn their place.
  6. When one success could reach further We take what works further.
  7. When results fall short of the promise We find the way back to value.

COGNIRISE / DECISION ROUTER

Start with your situation.

Select the decision in front of you. Find a practical next move, the evidence to bring, and methods that fit the question—not a prescribed sequence.

Cognirise methodology route map

Situation: We need to know where AI is worth investing.

Description: We are exploring AI, but have not established a useful opportunity or a convincing business case. We need to understand the likely benefits, costs and evidence before committing.

Existing assets: Strategic goals, cost pressures, baseline information or a few ideas may exist; a chosen project is not required.

Decision in front of you

What problem, if any, is worth addressing with AI, and what evidence would justify the investment?

Actual output: A bounded value hypothesis; comparison with non-AI alternatives; benefits, full costs, assumptions, owner and measurement plan; a proportionate next test or a reason not to proceed. This self-service route is not a complete business case; producing a full business case is additional engagement work.

Relevant methods: IDAO opportunity framing; AI Use-Case Prioritization

Why these methods: IDAO opportunity framing: Use Innovate when the opportunity, accountable owner, measurable outcome or proof plan is still unresolved. AI Use-Case Prioritization: Use only when actual comparable alternatives already exist; its scorecard informs sequence, not a complete financial business case.

Conditional IDAO connection: IDAO's canonical Innovate, Demonstrate, Activate and Operate gates remain evidence-dependent. Enter Innovate when the opportunity or evidence is unresolved; move to Demonstrate for a bounded hypothesis that needs proof, while adequate existing evidence can support a later entry.

Connection to Agent Authority: Use Agent Authority only when a specific consequential Knowledge, Decision or Action handover needs a ceiling. It is not a prerequisite for deciding whether to invest.

Situation: We have several AI ideas and need to choose.

Description: We have competing proposals and limited money or capacity. We need to decide what to fund first, what to investigate and what to stop.

Existing assets: Candidate descriptions, value hypotheses and some feasibility information exist, even if the evidence quality varies.

Decision in front of you

Which proposals should advance, in what order, and which should wait or stop?

Actual output: A transparent comparative decision with evidence gaps, capacity and control considerations, and investigate, demonstrate, activate or stop recommendations.

Relevant methods: AI Use-Case Prioritization; IDAO delivery gates

Why these methods: AI Use-Case Prioritization: Compare proposals across value, feasibility, time to evidence, adoption friction, control burden and reuse potential. IDAO delivery gates: Treat each recommendation as input to the canonical evidence gate; a score is not permission to bypass IDAO.

Conditional IDAO connection: Route each selected proposal according to its evidence: Innovate for unresolved opportunity, Demonstrate for a bounded test, Activate when implementation evidence is sufficient, or Stop when it should not proceed.

Connection to Agent Authority: Control burden and required oversight can change priority or the responsible IDAO entry. Use Agent Authority separately for each consequential handover.

Situation: We have an AI strategy and need to implement it.

Description: We already have a strategy or roadmap, whether developed internally or by a consultant. We need to turn it into owned, funded delivery work without repeating decisions that are already supported.

Existing assets: The strategy may include priority use cases, architecture, recommendations and a governance model; its completeness and evidence still need to be checked.

Decision in front of you

What can we start delivering now, who owns it, and what evidence or dependencies are still missing?

Actual output: An agreed first delivery package, accountable owner, funding and resourcing decisions, acceptance measures, dependencies and evidence-based IDAO entry. Turning a strategy or roadmap into delivery is additional engagement work beyond this self-service route.

Relevant methods: Direct IDAO entry; Supporting methods when a question remains

Why these methods: Direct IDAO entry: Review existing evidence and enter the earliest responsible gate; a roadmap does not require a strategy restart or bypass an evidence gate. Supporting methods when a question remains: Use prioritization, Value-to-Scale, readiness or work-design methods only for a specific unresolved decision—not a generic implementation score.

Conditional IDAO connection: Enter the earliest IDAO stage whose gate the existing evidence can responsibly satisfy. Demonstrate, Activate or Operate may be appropriate; return to Innovate only where the opportunity itself needs reframing.

Connection to Agent Authority: Review a specific consequential handover when one is part of the delivery package. Agent Authority does not replace the delivery gate; guardrails remain a complementary control question.

Situation: We need to improve a specific process.

Description: We know where work is slow, costly or unreliable. We need to work out whether AI would help and what must change in the process—not start by assuming it needs agents.

Existing assets: A known operational problem, affected work and ideally a process owner and baseline are available; an AI solution is not yet assumed.

Decision in front of you

What change would improve this process, and does AI have a useful role?

Actual output: A clear problem and baseline, changed-work proposal, AI and non-AI options, feasibility questions and a bounded test.

Relevant methods: IDAO opportunity framing; Human–Agent Operating Model; Agentic Operations Readiness

Why these methods: IDAO opportunity framing: Frame the bounded problem, evidence and next test without presuming an agent solution. Human–Agent Operating Model: Use when people, responsibilities, capabilities, incentives or handovers must change in the proposed work. Agentic Operations Readiness: Use only for an actual proposed agent workflow; it tests operating conditions, not every process-improvement option.

Conditional IDAO connection: Use Innovate for an unresolved intervention and Demonstrate for a bounded test. Do not require an agent-specific assessment when a non-agent process change is the better option.

Connection to Agent Authority: If the proposed change includes a consequential handover, define its authority separately. Guardrails and authority are specialist considerations, not an eighth starting situation.

Situation: We have a pilot and need to put it into everyday use.

Description: We have tested something, but it is not yet a supported part of normal work. We need to establish what remains before people can rely on it.

Existing assets: A prototype or limited trial exists, with some test evidence; integration, security, support, ownership, adoption or a safe release decision may remain unresolved.

Decision in front of you

What must be proven or completed before this can become a dependable part of normal work?

Actual output: An explicit proceed, prepare or stop decision where the agent-readiness method applies, or an IDAO release plan covering acceptance, integration, monitoring, fallback, support and people.

Relevant methods: IDAO delivery gate; Agentic Operations Readiness when agent-based; Human–Agent Operating Model when work changes

Why these methods: IDAO delivery gate: Use the canonical gate for a governed implementation and move into Operate only with the required release and ownership evidence. Agentic Operations Readiness when agent-based: Test the six operating conditions for the actual agent workflow; a promising demo does not automatically earn release. Human–Agent Operating Model when work changes: Design responsibilities, decision rights, enablement and adoption where everyday work will change.

Conditional IDAO connection: Remain in Demonstrate while evidence is insufficient; enter Activate when governed implementation is justified, then Operate once release, ownership and support evidence are in place.

Connection to Agent Authority: Set authority for each consequential handover before it operates. The approved guardrails-versus-authority distinction helps separate control requirements from the authority ceiling.

Situation: AI works in one area. We need to expand it.

Description: We have evidence of value in an existing live setting. We need to decide what can be reused and what must change for other teams, locations or workloads.

Existing assets: An operating use case has a baseline, demonstrated benefit, an owner and service or control experience; the new context may differ in data, language, permissions or accountability.

Decision in front of you

What can we reuse, what changes in the new context, and where should expansion proceed?

Actual output: An expansion decision and prioritized dependencies, with validated reuse assumptions, people and control changes, and renewed evidence where needed.

Relevant methods: IDAO Operate and loopback; AI Value-to-Scale; AI Use-Case Prioritization; Human–Agent Operating Model

Why these methods: IDAO Operate and loopback: Use operating evidence to inform expansion; new contexts may need Demonstrate or Activate rather than a blanket scale approval. AI Value-to-Scale: Use for repeated organizational constraints and evidence gaps across teams or locations, not as a financial business-case calculator. AI Use-Case Prioritization: Compare expansion choices when there are several candidate teams, contexts or workloads. Human–Agent Operating Model: Use where changed roles, capability, incentives, adoption or handovers determine whether reuse will hold.

Conditional IDAO connection: Operate evidence can inform expansion. Enter Demonstrate or Activate again where the new context changes the evidence; materially new opportunities may return to Innovate.

Connection to Agent Authority: Revisit each consequential handover when exposure, scope, evidence or operating context changes. Guardrails remain complementary to the canonical authority decision.

Situation: Our AI is in use, but the results are falling short.

Description: AI is already part of the work, but the benefits, quality, cost or adoption are disappointing. We need to identify the cause and decide whether to improve, redesign, replace or stop it.

Existing assets: A live deployment and some usage or performance evidence exist; a measurement gap must be acknowledged rather than filled with an assumed result.

Decision in front of you

Why are the results insufficient, and should we improve, redesign, replace, reduce scope or stop?

Actual output: A diagnosis against an agreed baseline and outcome, a targeted corrective decision with an owner, and a retest or stop criterion.

Relevant methods: IDAO operating evidence and loopback; Human–Agent Operating Model; AI Value-to-Scale; Agentic Operations Readiness when conditions fail

Why these methods: IDAO operating evidence and loopback: Use live evidence to improve, constrain or return to the responsible earlier decision; do not automatically prescribe another pilot. Human–Agent Operating Model: Use directly when responsibilities, incentives, capabilities, adoption or handovers are the known cause. AI Value-to-Scale: Use for repeated systemic constraints across the organization, not as a universal failure score. Agentic Operations Readiness when conditions fail: Use for an unsafe or underperforming agent workflow to identify unresolved operating conditions.

Conditional IDAO connection: Stay in Operate for bounded improvement when the proposition remains sound; return to Demonstrate or Innovate when assumptions fail. Stop or constrain unsafe or unviable work.

Connection to Agent Authority: If a known problem concerns a consequential handover, use Agent Authority directly and reassess when exposure, scope or evidence changes. Review the existing guardrails distinction alongside it.

Direct specialist routes

Human–Agent Operating Model: Open the playbook directly when changed work, roles, capabilities, adoption or handovers are already the decision.

Agent Authority Model: Calculate the ceiling directly for a specific consequential Knowledge, Decision or Action handover.

AI Guardrails: Planned as a complementary specialist method alongside Agent Authority. No standalone Guardrails framework or assessment is published; read the approved guardrails-versus-authority distinction in the existing Agent Authority page.

Open Human–Agent Operating Model

Open Agent Authority Model

Read approved guardrails-versus-authority distinction

ONE CONTINUOUS VIEW

Choose a threshold

01 / 07

Seven real starting points · one practical next decision

01 / 07 · CURRENT STOPWe need to know where AI is worth investing.
Selected situation: We need to know where AI is worth investing.

We need to know where AI is worth investing.

We are exploring AI, but have not established a useful opportunity or a convincing business case. We need to understand the likely benefits, costs and evidence before committing.

Existing assets

Strategic goals, cost pressures, baseline information or a few ideas may exist; a chosen project is not required.

Decision needed

What problem, if any, is worth addressing with AI, and what evidence would justify the investment?

Actual output

A bounded value hypothesis; comparison with non-AI alternatives; benefits, full costs, assumptions, owner and measurement plan; a proportionate next test or a reason not to proceed. This self-service route is not a complete business case; producing a full business case is additional engagement work.

Relevant methods and why they fit
IDAO opportunity framing

Use Innovate when the opportunity, accountable owner, measurable outcome or proof plan is still unresolved.

AI Use-Case Prioritization

Use only when actual comparable alternatives already exist; its scorecard informs sequence, not a complete financial business case.

Connections to Governing Frameworks
IDAO Delivery Framework
Conditional IDAO connection

IDAO's canonical Innovate, Demonstrate, Activate and Operate gates remain evidence-dependent. Enter Innovate when the opportunity or evidence is unresolved; move to Demonstrate for a bounded hypothesis that needs proof, while adequate existing evidence can support a later entry.

Agent Authority Model

Use Agent Authority only when a specific consequential Knowledge, Decision or Action handover needs a ceiling. It is not a prerequisite for deciding whether to invest.

Direct specialist access

Already know the specialist decision?

These are direct routes, not additional starting situations or mandatory steps.

Human–Agent Operating Model

Go directly here when changed work, roles, capabilities, adoption or handovers are the known decision.

Agent Authority Model

Calculate the ceiling directly for a specific consequential Knowledge, Decision or Action handover.

AI Guardrails: planned as a complementary specialist method alongside Agent Authority. No standalone Guardrails framework or assessment is published yet. Read the approved guardrails-versus-authority distinction in the existing Agent Authority page.

Value Scan remains a separate optional facilitated enquiry, not a scored assessment. IDAO and the Human–Agent Operating Model remain delivery and playbook content; they are not presented here as compulsory questionnaires.