Home / Operating Rhythms
STRATEGY, OPERATIONS & CONTINUOUS LEARNING
AI transformation rhythms.
Connect evidence to decisions.
AIDIVE operating rhythms connect AI strategy, delivery and operations through evidence and accountable decisions. Explore 31 V1 rhythms and adapt their purpose, ownership and outputs to the way your enterprise works.
TAKE THE NEXT STEP
A cadence is more than
a recurring meeting.
Operating rhythms connect current evidence to accountable decisions and learning. Integrate them into existing enterprise and delivery cadences. A cadence states when an interaction is needed; it does not prescribe a separate meeting.
Decision or learning purpose
Accountable authority
Required evidence
Frequency or trigger
Recorded output
Escalation path
The V1 rhythm library
AR-R01 · Data
Weekly Data Readiness Triage
Resolve new blockers and quality/access exceptions affecting priority use cases.
Evidence: New issues, threshold breaches, dependency changes and decisions due.
Output: Owned actions, safe-state decision, escalation or closure evidence.
V1 cadence: Weekly during delivery, embedded in existing flow triage
AR-R02 · Data
Biweekly Context & Retrieval Evaluation
Review evaluation regressions, authority/freshness failures, permission tests and context backlog.
Evidence: Evaluation scorecard, failed cases, corpus/index changes and release candidate.
Output: Approved fixes, updated evaluation set and release conditions.
V1 cadence: Biweekly during build and material change, embedded in existing evaluation cadence
AR-R03 · Data
Release Data & Context Gate
Confirm lifecycle boundaries, data contracts, permissions, reproducibility, lineage and operational readiness.
Evidence: Release evidence pack: lifecycle boundaries, interface and tool contracts, permission tests, reproducibility evidence, lineage completeness and operational readiness results, together with open exceptions.
Output: Release, conditional release with expiry, or no-go decision.
V1 cadence: At each material release
AR-R04 · Data
Monthly Data Trust Conversation
Connect data and context signals to AI behaviour, incidents, overrides and business outcomes.
Evidence: Trust dashboard, production retrieval sampling, incidents, remediation status and KPI movement.
Output: Risk treatment, remediation priorities and control changes.
V1 cadence: Monthly in production, embedded in existing operating-health cadence
AR-R05 · Data
Quarterly Dependency & Portfolio Risk Conversation
Review shared data dependencies, concentration risk, ownership, capability investment and roadmap outcomes.
Evidence: Cross-use-case dependency view assembled from the critical asset register and impact analyses, risk register, benefits and capacity.
Output: Portfolio priorities, shared-capability investment and accepted residual risk.
V1 cadence: Quarterly, embedded in existing portfolio and risk cadence
AR-R06 · Alignment
Weekly Alignment Flow & Blocker Triage
Resolve ownership, dependency, capacity and decision blockers that prevent the next evidence step.
Evidence: Ageing and blocked work; WIP limits; decisions due; dependency and readiness changes; unresolved owner actions.
Output: Owned unblock action, delegated decision, escalation with evidence, WIP adjustment or stop/pause decision.
V1 cadence: Weekly during active delivery, embedded in existing delivery flow
AR-R07 · Alignment
Biweekly Hypothesis & Evidence Conversation
Review assumptions and evidence, then decide whether to continue, pivot, add evidence, pause or stop.
Evidence: Hypothesis cards; experiment results; user/adoption evidence; data/model findings; risk signals; costs and next learning need.
Output: Evidence decision, updated hypothesis, next smallest test, changed scope or portfolio-state decision.
V1 cadence: Biweekly during Explore and Validate, embedded in existing learning and evidence cadence
AR-R08 · Alignment
Monthly AI Portfolio Flow & WIP Conversation
Balance investment, capability demand and change capacity while managing WIP, ageing and portfolio mix.
Evidence: Portfolio Kanban; ageing and flow data; prioritisation evidence; shared dependencies; capacity; benefits, adoption and risk movement.
Output: Start/wait/enable/scale/pause/redesign/stop decisions, WIP and capacity changes, shared-capability investment.
V1 cadence: Monthly within existing portfolio decision cadence
AR-R09 · Alignment
Monthly Decision & Governance Health Conversation
Improve decision speed and accountability by reviewing latency, reversals, escalations, exceptions and evidence quality.
Evidence: Decision log; latency and ageing; escalations; exceptions; DACI gaps; decision-mode changes; incidents and follow-through.
Output: Decision-right changes, delegated authority, forum changes, evidence-contract fixes and overdue action resolution.
V1 cadence: Monthly within existing governance decision cadence
AR-R10 · Alignment
Quarterly Strategy, OKR & Roadmap Reframing
Reframe strategic themes, OKRs, portfolio balance and roadmap investment from accumulated evidence.
Evidence: Outcome and OKR movement; hypothesis learning; portfolio flow; benefits; adoption; operational and risk signals; market/regulatory changes.
Output: Confirmed or revised strategic intent, themes, guardrails, OKRs, investment and capability priorities.
V1 cadence: Quarterly, aligned to enterprise planning
AR-R11 · Alignment
Trigger-Based Alignment Reset
Restore coherent intent, authority and communication after a material strategy, risk, evidence or change event.
Evidence: Trigger and impact assessment; affected hypotheses, decisions and stakeholders; current narrative; safe-state and open risks.
Output: Updated direction, guardrails, decision rights, roadmap state, stakeholder narrative and verification actions.
V1 cadence: Within the decision time defined for a material trigger
AR-R12 · Integration
Weekly Integration Flow Triage
Resolve blockers, handoff failures, orphan outputs, context gaps and unresolved execution issues affecting priority AI-enabled flows.
Evidence: New flow exceptions and handoff failures; queue and latency signals; orphan outputs; unresolved decisions; context-loss cases; actions due.
Output: Owned actions, escalation or closure evidence; an explicit safe-state decision where the issue affects a critical flow.
V1 cadence: Weekly during delivery and stabilization, within existing delivery or operations flow
AR-R13 · Integration
Biweekly Autonomy and Control Conversation
Review AI authority mode, human-control mode, control activations, overrides, permission or action-envelope changes, escalations and unresolved control conditions.
Evidence: Control-activation logs; override and escalation cases; AI authority or human-control mode changes; permission and action-envelope changes; failed interactions; unresolved control conditions.
Output: Retain the current operating design; approve or condition a change within delegated authority; request additional evidence; assign remediation; narrow or hold authority; or escalate the decision to the required authority.
V1 cadence: Biweekly during build, change and early production, embedded in existing control and decision cadence
AR-R14 · Integration
Release and Exposure Expansion Gate
Decide whether the proposed release or exposure transition has sufficient Integration evidence to proceed, proceed with conditions, be held, narrowed or reversed.
Evidence: Change-specific evidence covering the affected flow, AI authority and action envelope, permissions and controls, interaction and execution tests, evaluation results, continuity and recovery, ownership and traceability, and open exceptions.
Output: Proceed; proceed with explicit conditions and owner; hold; narrow the proposed exposure; or reverse the affected transition, with the next evidence requirement and decision authority recorded.
V1 cadence: At each material release or proposed increase in AI authority, population, business consequence or operating exposure
AR-R15 · Integration
Monthly Operational Integration Health Conversation
Review conversion, latency, evaluation performance, control and exception activity, support load, incident signals and continuity tolerance across material AI-enabled flows.
Evidence: End-to-end conversion and latency; evaluation results; control interventions and exceptions; execution failures; incidents; support demand; continuity or degraded-operation events; overdue operational actions.
Output: Prioritised remediation, accepted residual risk, change decision and verified closure criteria.
V1 cadence: Monthly in production, embedded in existing operational-health cadence
AR-R16 · Integration
Quarterly Scale, Debt and Resilience Conversation
Review shared dependencies, concentration, resilience evidence, estate exposure, portability, debt carrying cost and unresolved lifecycle risk before further scale or replication.
Evidence: Integration estate and dependency views; debt and dependency register; resilience and recovery-test results; concentration exposure; portability and exit findings; lifecycle status; carrying-cost and remediation evidence.
Output: Scale or replicate; hold or narrow expansion; accept, reduce or resolve debt; invest in resilience; consolidate or replace a dependency; or initiate an exit action, with accountable owner and review condition.
V1 cadence: Quarterly and before material replication or operating-exposure expansion, embedded in existing portfolio and resilience cadence
A named cadence does not replace formal authority. Confirm the owner, evidence, output and escalation route for the actual decision context.
LEARN. CONTRIBUTE. APPLY.
Put AI transformation
into practice.
Build capability through training, contribute your expertise or explore a connected workspace for applying the holistic AIDIVE framework.
TRAINING & CERTIFICATION
Build the capability.
Develop the skills to connect AI transformation strategy, business outcomes, people and governance to your own decisions and responsibilities.
PARTNER WITH AIDIVE
Grow the practice.
Bring your learning, delivery or research expertise to the holistic AIDIVE approach to enterprise AI transformation.
AIDIVE PLATFORM
Bring AIDIVE into your work.
See how the proposed workspace could connect AI transformation learning, maturity evidence and practical artifacts.
CONNECT THE DOTS
There is more to explore.
AR-R17 · Values
Weekly Champion Office Hours
Resolve local questions quickly, surface weak signals and keep champions connected to central expertise. This is a capability-support mechanism, not a governance forum.
Evidence: Review the three highest-impact open questions.; Triage: answer, pair, escalate or add to the learning backlog.; Capture one reusable pattern or clarification.; Confirm owners and response dates.
Output: Triage log; escalations; reusable answer or case; learning backlog update.
V1 cadence: Weekly, through existing champion support and office-hours channels
AR-R19 · Values
Leadership Values Lens
Model evidence-based questioning, make uncertainty discussable and close executive-to-middle-manager learning loops.
Evidence: Review one consequential decision and its evidence.; Examine confidence, missingness, human judgement, approval behavior and affected stakeholders.; Review one concern, near miss or workforce friction, including where enthusiasm and uncertainty coexist, and the organizational response.; Agree one leadership or manager-system change, owner and closure date.
Output: Decision record; unknowns/evidence backlog; manager escalation closure; leadership learning note.
V1 cadence: embedded in the existing executive strategy and alignment conversation
AR-R21 · Values
Value Perception and Evidence-Confidence Lens
Maintain calibrated confidence and credible value perception without duplicating Impact measurement or Alignment portfolio governance.
Evidence: Reference the Impact-owned outcome evidence and its stated limits.; Observe how confidence and stakeholder perception compare with the formal evidence.; After the owning forum makes its stop, pivot, continue or scale decision, test comprehension, trust and responsible dissent.; Update perception gaps, narrative-integrity records and cross-dimension referrals.
Output: Link to Impact evidence; decision-trust pulse; perception/confidence log; narrative-integrity record; cross-dimension referrals.
V1 cadence: embedded in the Portfolio Value & Evidence Conversation
AR-R23 · Ethics
Biweekly Lifecycle Risk and Change Conversation
Review new intakes, tier changes, material changes, lifecycle evidence and conditional approvals across the portfolio.
Evidence: Inventory delta; gate pipeline; material-change triggers; monitoring breaches; evidence gaps; vendor changes.
Output: Risk-tier and assurance recommendations; reassessment assignments; expired exceptions; unresolved portfolio risk and evidence backlog.
V1 cadence: Biweekly in existing lifecycle risk and change governance
AR-R25 · Ethics
Monthly Incident, Recourse and Learning Conversation
Convert incidents, near misses, complaints, overturns and remedies into changed controls and operational learning.
Evidence: Incidents and near misses; recourse cases; root causes; remedy status; repeat patterns; control and training actions.
Output: Systemic cause decisions; control/policy updates; remedy closure; disclosed lessons; residual-risk escalation.
V1 cadence: Monthly in existing incident, recourse and learning mechanisms
AR-R27 · Impact
Weekly Impact Evidence Triage
Resolve evidence defects, assign owners and dates, escalate threshold breaches, and keep the next decision’s evidence pack complete.
Evidence: New outcome and guardrail signals; missing or late evidence; metric-quality defects; experiment status; trigger breaches; decisions due in the next two weeks.
Output: Resolve evidence defects, assign owners and dates, escalate threshold breaches, and keep the next decision’s evidence pack complete.
V1 cadence: Weekly during discovery, delivery and early operation; event-driven for critical evidence or guardrail issues, within existing evidence and delivery cadence.
AR-R29 · Impact
Initiative Evidence and Decision Gate
One recorded decision proceed, investigate, pivot, pause, scale, reduce autonomy or retire with conditions, owners, evidence obligations and next date.
Evidence: Value intent; outcome-chain status; outcome and guardrail evidence; fully loaded economics; confidence and counterfactual; ownership; unresolved risks; trigger history.
Output: One recorded decision proceed, investigate, pivot, pause, scale, reduce autonomy or retire with conditions, owners, evidence obligations and next date.
V1 cadence: At every material funding, release, autonomy, scale, pivot or retirement decision.
AR-R31 · Impact
Semi-Annual Scale and Sustainment Conversation
Confirm sustainment investment, redesign the scale envelope, remediate local value leakage, switch providers or models, or retire value-eroding scope.
Evidence: Value health by site and segment; dilution; portability; service and incident trends; unit economics; capability dependencies; contract and exit exposure; sunset criteria.
Output: Confirm sustainment investment, redesign the scale envelope, remediate local value leakage, switch providers or models, or retire value-eroding scope.
V1 cadence: Semi-annually, or upon material model, vendor, regulatory, demand or operating-model change, embedded in sustainment and portfolio decisions.
AR-R18 · Values
Biweekly Community Learning Loop
Convert local experience into reusable capability and reduce dependency on individual experts.
Evidence: One real case, not a presentation deck.; What was expected, observed and learnt?; Pair to create or improve one reusable asset.; Select the next case and rotate facilitation.
Output: Validated pattern or case card; owner/review date; dependency and community backlog update. Reusable outputs are routed to the relevant team or portfolio backlog and review.
V1 cadence: Biweekly, through existing community learning channels
AR-R20 · Values
Human–AI Work Values Lens
Inspect total workflow experience, workforce transition, learning conditions, open AI use and agent-supervision load after AI changes.
Evidence: Review the Impact-owned outcome evidence together with employee experience and work-design observations.; Surface new checking, accept/edit/reject/escalate patterns, hidden tool use, learning barriers, early-career or tradecraft concerns, agent-to-agent handoffs and supervision burden.; Investigate possible automation bias, approval fatigue, unsustainable supervision span or loss of manual recovery capability.; Select one friction to remove, one transition or capability commitment to close and one human-experience guardrail to protect.; Record the Values recommendation and route any formal work-design, autonomy or portfolio decision through the owning A&R mechanism.
Output: Work/friction backlog; oversight-effectiveness finding; workforce, early-career or tradecraft commitment; Values recommendation; role/guidance change; employee closure note.
V1 cadence: embedded in the Human-Agent Work Conversation
AR-R22 · Ethics
Weekly Ethics Control Triage
Resolve urgent policy, guardrail, disclosure and agent-action questions; contain emerging risk and keep delivery moving.
Evidence: Highest-consequence open questions; blocked actions; exceptions; complaints; near misses; upcoming release gates.
Output: Triage decisions; named owner/date; containment or escalation; reusable pattern update.
V1 cadence: Weekly in existing control/delivery triage; immediate escalation for material triggers
AR-R24 · Ethics
Monthly Agentic-AI Oversight Conversation
Inspect authority, identity, tools, trajectories, supervision load, incidents and evidence supporting proposed changes in agent autonomy or delegated authority.
Evidence: Agent inventory; authority envelopes; permission changes; action/approval traces; evals; blocked actions; supervisor load.
Output: Ethics assurance findings and recommendations on proposed autonomy or delegated-authority changes; permission-reduction or containment actions; new tests; supervision-capacity and retirement actions; authority changes routed to the approved decision owner.
V1 cadence: Monthly in existing agentic oversight and runtime governance
AR-R26 · Ethics
Quarterly Responsible Scale and Regulatory Conversation
Provide Ethics assurance for proposed scale or restriction decisions and set assurance priorities using current obligations, evidence, claims and cumulative risk.
Evidence: Scale proposals; obligations horizon; assurance cases; external claims; audit results; supplier concentration; systemic incidents.
Output: Ethics assurance recommendation for scale/restriction; updated obligations/control roadmap; claim corrections; independent-review plan.
V1 cadence: Quarterly in existing regulatory and scale governance, plus material-change triggers
AR-R28 · Impact
Monthly AI Economics and Benefit Conversation
Approve forecast updates, cost and workflow interventions, benefit recognition, trigger actions and any change to the operating or model mix.
Evidence: Token and runtime cost; fully loaded cost per usable outcome; volume, adoption and goodput; realised benefit; forecast variance; benefit-realisation actions; anomalies.
Output: Approve forecast updates, cost and workflow interventions, benefit recognition, trigger actions and any change to the operating or model mix.
V1 cadence: Monthly for production initiatives, with event-driven action during rapid volume, price or model change, within existing financial and product cadence.
AR-R30 · Impact
Quarterly Portfolio Value Conversation
Reallocate capital and scarce capacity, reconcile shared value, set evidence obligations and change the portfolio’s fund/test/hold/stop mix.
Evidence: Portfolio balance by intent and evidence stage; confidence-calibrated ROAI; shared enablers; dependencies; concentration; realised benefit; stop/reallocate candidates.
Output: Reallocate capital and scarce capacity, reconcile shared value, set evidence obligations and change the portfolio’s fund/test/hold/stop mix.
V1 cadence: Quarterly, aligned with the existing capital and capacity allocation cycle.