VisionVolve· Engage

Methodology summary · v0.4 · 2026-09-02

AI Transformation at Scale — Engagement Methodology

Instantiated for: Dr. Max — Engaged Target (M6)

VV's deep-engagement offering: a time-boxed, fractional, board-mandated Interim AI Transformation Partner role that installs a complete AI operating system in ~6 months and exits, leaving behind durable internal AI leadership, governance, strategy, policies, processes, a working PoC pipeline, and trained people.

1. The methodology in one paragraph

What we install, how we measure it, when we leave

Three lenses (process / system / logic) instantiated as six parallel tracks. A connective BDN layer ties initiatives, capabilities, benefits, and objectives. A META charter (CAIO Operating Charter Q1–Q8) defines how the Partner role itself operates. The framework is evidence-grounded: anchored on Cranfield/Ward & Peppard BDN, Gartner AI Maturity, NIST AI RMF, ISO/IEC 42001, EU AI Act, Cooper Stage-Gate.

Engagement archetype

Interim AI Transformation Partner — fractional (typically 2/5 days/week), time-boxed (typically ~6 months), board-mandated, exit-oriented. Not advisory. Not permanent. Authority-backed. Recruits VV delivery layer separately; preserves clean commercial interface.

Reading guide

The glossary in §2 below decodes forum names (Portfolio Council, AI Council, Scale Decision Forum), gate names (Heavy Gate 1/2), BDN levels, and engagement phases. Skim it once before the rest — every later section assumes these names.

2. Glossary · names you'll see

Decode forums, gates, levels, phases

Methodology vocabulary that appears throughout this document and on every Engage artefact (Charter Q1–Q8, M0 brief, dashboards). Each term is a load-bearing concept, not jargon for its own sake — the glossary entry explains what it does and why it exists.

Portfolio Council: Recurring forum (typically bi-weekly or monthly) where AI initiatives are reviewed, ranked, advanced, killed, or parked. The body that makes Heavy Gate 1 decisions. Track 1 (Pipeline) backbone.
AI Council: Central governance forum (typically monthly) where AI strategy, risk, vendor, and budget decisions converge. Track 3 (AI OS) backbone — the operating system that runs everything else.
Scale Decision Forum: Body that makes Heavy Gate 2 decisions (Pilot-Validated → Scaling). Different members + cadence from Portfolio Council because capex + operational change are at stake. CFO + COO + CRO must be present.
Heavy Gate 1: Formal Sponsorship → Pilot decision. The point where pilot capacity gets committed. Below it, cheap experiments. Above it, real resources allocated.
Heavy Gate 2: Formal Pilot-Validated → Scaling decision involving capex commit. Where AI investment becomes real money. Failure mode: pilots accumulate without crossing this gate.
4-Gate framework: VV-original early-phase evaluation framework: WHY → HOW → WHAT → WHERE/WHEN. Run at Scoped stage to test strategic fit, archetype, build feasibility, and timing before committing pilot resources.
BDN (Benefit Dependency Network): Directed graph linking Drivers → Objectives → Benefits ← Changes ← Foundations (with Risk Constraints overlay). Methodology v0.4 §5. Three snapshot levels: Initiative-BDN, Portfolio-BDN, Strategy-BDN.
Initiative-BDN: BDN snapshot for one initiative. Versioned: v0 at Scoped, v1 at Pilot-Validated, v2/v3 in Realizing. The delta v3 − v0 is the organizational learning artifact.
Portfolio-BDN: Aggregate BDN across all active initiatives, with shared nodes merged. Quarterly cadence. Where critical-path detection (A2 analysis) and change-burden (A3) live.
Strategy-BDN: Highest-level BDN: board investment objectives → portfolio composition. Annual + on strategy events. Drives reallocation conversations.
Internal AI Lead (IAL): Client-side counterpart to the Partner. Title varies (CAIO / Head of AI / VP AI / AI Director). Always elevated to senior + mandated by engagement end. For Dr Max: Kateřina Lesch.
Interim AI Transformation Partner: The fractional, time-boxed, board-mandated executive role VV provides. For Dr Max: Michal Lickó. ~6 months, ~25% FTE. Authority-backed, not advisory.
Day 0 / PE / M0 / M1-M6 / Advisory tail: Engagement phases. Day 0 = pre-engagement state captured but not contracted. PE = pre-engagement workstream (PE-1 to PE-5). M0 = configuration meeting where charter signs. M1-M6 = main 6-month engagement. Advisory tail = optional post-engagement support.
Track maturity index: Composite score across the 6 tracks at a given checkpoint. Q5 Partner KPI: monotonically increasing throughout engagement.
% stable-without-us: Per-track score 0-100 estimating how much the track would continue functioning if the Partner left tomorrow. Leading indicator of exit readiness. Target: ≥80% on Track 1, ≥70% on Track 2 by M6.
Healthy kill rate: Q5 Partner KPI. Target: ≥30% of initiatives killed before Pilot stage. Why: a too-low kill rate means Triage discipline is broken — bad ideas advancing.

3. Five design invariants

Non-negotiable principles

Every engagement must satisfy these. Stress-tested across all archetypes (Greenfield / Elevation / Realignment / Turnaround) and overlay dimensions (multi-market, regulatory profile).

Real authority, fractional presence

Board access + finance data access + decision rights on AI portfolio. Without explicit mandate, Partner becomes another expensive advisor.

System-setter, not daily decider

Partner designs mechanisms others run. Day-to-day decisions happen in the forums Partner stood up.

Successor-building from Day 1

Internal AI Lead identification (or elevation) is Week 1. Every Partner action prepares the successor (R4).

Exit criteria agreed upfront

Contract names "done." Q6 of the META charter is signed at M0, not negotiated at exit.

Clean commercial interface

Partner cannot approve VV's own scope expansions. Recusal protocol with named SLA. Three-layer fee structure (retainer + delivery + specialists) keeps trails independent.

4. Three-lens model (the framework's philosophy)

Funnels, stock & flow, BDN — three views of the same world

The methodology reasons about engagements through three complementary lenses. They are not redundant — each answers a different question and operates on a different object.

Funnels

PROCESS · per-item state

Object: each individual thing (initiative, person, organization).

Question: Where does this one sit, and what moves it next?

Blind to: aggregate dynamics.

Stock & Flow

SYSTEM · populations + rates

Object: populations of things at scale.

Question: Are we accumulating enough of the right things, fast enough?

Blind to: item identity.

BDN

LOGIC · causal chains

Object: causal chains between things.

Question: If this changes, what else changes?

Blind to: time / process state.

The mental model

Same engagement, three views. BDN is the logic layer (snapshots at gates; mostly stable between snapshots). Funnels are the process layer (updated continuously as work happens). Stock & flow is the system layer (computed from funnel data; signals checked against BDN dependencies).

5. Six tracks · the operational dimensions

Six parallel dimensions of progress

Each track instantiates the three lenses to a specific domain. Tracks are not silos — six explicit cross-track couplings (§6 below) make engagement health a system property.

T1 Opportunity Pipeline funnel · funnel

Initiative motion from idea capture to value realization. Two-phase split: cheap experiments + heavy gates.

Current state

Stage 1–2 (Captured exists, no flow) Drain existing initiative list through 4-Gate Week 1

Target at exit

≥ 80% stable-without-us 2 cycles of gates without Partner intervention

T2 People stock-and-flow · metro-map

Capability stocks across four paths. Stocks bound flows in Track 1.

Current state

A1 thin, B0–1 embryonic, C partial, D mixed Group + per-market champion mapping needed

Target at exit

A2 ≥60%; B-stock ≥ pilot demand; D2 100% Champion induction + sponsor pairing operational

T3 AI Operating System BDN · maturity-ladder

The governance backbone that runs everything else. Without this, tracks 1, 2, 4, 5, 6 are uncoordinated.

Current state

Level 0–1 (Absent / Chartered) Group-vs-market authority must be settled first

Target at exit

Level ≥ 5 (Self-running) Internal AI Lead chairs without Partner; ≥ 2 cycles autonomous

T4 Value Chain Coverage funnel · maturity-ladder

How much of the business value chain has AI initiatives mapped against it. Without coverage, pipeline starves at source.

Current state

Level 0–1 (Reactive / Anchor mapped) Multi-market: group-level Level 3 target + per-market Level 1–2

Target at exit

Level ≥ 3 (group); Level 1–2 per-market acceptable Depth achieved on group; market customization allowed

T5 Risk & Compliance BDN · maturity-ladder

Defensible AI risk posture. Mandatory work; not optional.

Current state

Level 1 (Classified — some classification likely exists) EU AI Act + GDPR + health + loyalty data → need Level 2 by mid-engagement

Target at exit

Level ≥ 4 (Monitored) High regulatory + multi-market = floor at L4; live monitoring + DPIA on file + ISO 42001 readiness pack

T6 Value & Reporting BDN · maturity-ladder

How AI value gets measured + reported. Track 6 is the connective layer between AI initiatives and the P&L.

Current state

Level 0 (Opaque) Board-pack rhythm earlier than default given Dr Max board engagement

Target at exit

Level ≥ 3 (P&L-linked); Level 4 strongly preferred CFO sign-off on benefits framework; board pack rhythm

Multi-market overlay (§2.5)

When the engagement spans multiple markets/jurisdictions, an independent overlay applies on top of the base archetype. Per-market peer expansion, multi-DPO coordination protocol, group vs market decision split. Highest-profile market sets the regulatory floor.

6. Cross-track couplings · the wiring

Why tracks aren't silos

Engagement health is a system property. A weak Track 2 → Track 1 throughput collapses regardless of Track 1 design quality.

From	To	Mechanism	What happens when broken
Track 4 →	Track 1	Map produces opportunities into Captured stage	Pipeline starves on inbound-only ideas; politics drives portfolio
Track 2 (B+D) →	Track 1	B-stocks bound Initiative Lead supply; D-stocks bound Sponsor supply	Pipeline stalls at Sponsorship Gate; pilots run un-led or un-sponsored
Track 5 →	Track 1	Risk classification gates Scoped; re-check gates Pilot→Scaling	High-risk initiatives slip through; regulatory exposure
Track 6 ⟲	Track 1	Realizing → portfolio reallocation → Sponsorship votes	Pipeline runs without consequence; theatre ensues
Track 3 →	all	Forums + decisions = the OS that runs everything else	Tracks operate as silos; coordination fails
Track 6 →	Track 3	Track 6 Level 3+ requires Track 3 Level 4 (CFO integration)	Reporting cannot bridge to financial language
BDN ⤳	Tracks 1, 6	Snapshots inform gate decisions and benefits review	Gate decisions become opinion-driven; benefits untraceable

7. BDN connective layer (§5)

The logic layer that ties tracks together

Six node types (Drivers / Objectives / Benefits / Changes / Foundations / Risk Constraints) in a directed dependency graph. Three snapshot levels: Initiative-BDN (per initiative; v0 at Scoped, v1 at Pilot-Validated, v2/v3 at Realizing), Portfolio-BDN (aggregate, quarterly), Strategy-BDN (board investment objectives → portfolio composition; annual).

Seven named BDN-enabled analyses

A1 Kill-impact — if we kill X, which benefits / objectives are hit? (at any kill decision)
A2 Critical-path — which foundations/changes are shared by ≥3 initiatives? (quarterly Portfolio-BDN review)
A3 Change-burden — which audience faces the most change requests across the portfolio? (quarterly + before adding new initiatives)
A4 Reality-check — are claimed benefits proportional to foundations actually delivered? (at Pilot-Validated and Realizing snapshots)
A5 Scarcity-event — capacity drops on resource X — which initiatives at risk? (on capacity event)
A6 Strategy-pivot — objective changes — re-rank initiatives by alignment to new objectives (on strategy event)
A7 Snapshot-delta — what was true at Scoped vs. Realizing — what did we learn? (at Realizing-Q4)

8. Pre-engagement workstream (§1.7)

The phase before the 6-month engagement

Distinct phase preceding the main engagement. Chargeable separately. Purpose: convert a verbal engagement commitment into a signed, executable Engagement Charter.

Step	Deliverable	Typical duration
PE-1	Stakeholder map + interview pass (IAL candidate, sponsor, board chair, key peer functions)	2–3 weeks
PE-2	Q1–Q8 META charter draft instantiated for the client	1 week
PE-3	Initial track baseline assessment using diagnostic matrices	2 weeks
PE-4	Obstacle Risk Profile run — which obstacles likely activate?	Concurrent with PE-3
PE-5	Engagement Charter signed; Day 1 calendar locked	1 week

Total typical pre-engagement: 6–8 weeks of part-time work (~30–40% of a Partner FTE during this phase). Multi-market or high-regulation engagements extend toward the upper bound.

9. META layer · CAIO Operating Charter (§8)

How the Partner role itself operates — Q1–Q8

The META layer sits above the operational tracks. It defines how the Partner role operates: mandate, authorities, decision rights, escalation, KPIs, exit criteria, commercial interface, peer interfaces. Without META, the framework describes a transformation but not the role that runs it.

Question	What it defines	Calibration window
Q1 Mandate scope	Group-wide unified / Methodology+standards central / HQ-focused / Hybrid	PE-2 (charter draft)
Q2 Authority set	Must-have / Nice-to-have / Out / Recused — by L1 ID	PE-2 + M0 negotiation
Q3 Decision rights	Decide / Propose / Recommend / Recused — per decision type	PE-2 + M0
Q4 Escalation	Bidirectional (R21) — Partner can escalate up; engagement can escalate Partner up. Recusal SLA.	PE-2; signed at M0
Q5 Partner KPIs	Operating-quality measures (track maturity index, healthy kill rate, decision register completeness)	PE-2; reviewed monthly
Q6 Exit criteria	Hard (8) + soft (5) criteria. Regulatory profile overrides Track 5 floor. Multi-market floor rule.	Signed at M0; non-negotiable thereafter
Q7 Commercial interface	Three-layer fee + pre-engagement + bonuses. Recusal protocol.	PE-2; signed at M0
Q8 Peer interfaces	Patterns + owned issues + conflict-prone areas per C-suite role. Multi-market expansion.	PE-2; reviewed quarterly

10. Engagement velocity model (§7)

No fixed phases — adaptive parallel tracks + 4-weekly checkpoints

Phase-based plans (discovery month → design quarter) don't fit clients arriving with prior failures. Replaced by parallel tracks moving from Day 1 + 4-weekly velocity checkpoints. For each track: Where does it sit on stages/levels? Delta vs. last checkpoint? Current blocker? % stable without us (0–100; leading indicator of exit readiness)? Risks/issues + escalation flag.

Plus quarterly Portfolio-BDN review (A2 critical-path + A3 change-burden + A4 reality-check).

11. Authority Erosion × Elevation (§2.4) — dominant pattern

The risk every Elevation engagement must counter actively

Critical for Elevation engagements

In Elevation engagements, the existing Internal AI Lead's authority is the load-bearing structure of the entire transformation. Every Partner action is a stress on that structure. Authority Erosion is not a risk that may activate — it is the dominant pattern requiring active counter-pressure throughout the engagement.

Mandatory counter-mechanisms (Elevation)

Visibility-elevation rituals — Lead chairs forums, Lead presents to board, Lead is the named decision-maker on AI portfolio
Partner-as-amplifier-not-replacer messaging — consistent across all client communications
Attribution discipline — "Lead is the decision-maker; Partner is the framework provider" stated explicitly in board updates and town halls
Recusal-equivalent on Lead's domain decisions — Partner advises but does not adjudicate decisions where Lead has clear authority
Lead self-rating in exit scorecard — Lead's own readiness assessment is a hard exit criterion (Q6)

12. Application read · track baselines (this engagement)

Where this engagement starts and where it ends

Per Methodology v0.4 §10 (Dr Max application read). At PE-3 the current state IS the typical baseline — comparing against generic week-12-typical would mis-fire.

13. Design invariants checklist (this engagement)

Pre-meeting check before M0

Methodology v0.4 · Generated 2026-09-02 · instantiated for Dr. Max — Engaged Target (M6)
Source: VisionVolve_AI_Transformation_at_Scale_Methodology_v0_4.md