Governance Model — PRIME PRODUCTS AI-First Transformation

This document defines how the program is steered, who decides what, how issues escalate, and how progress is reported. It is binding on all workstreams of the project charter. Design principle: the lightest structure that keeps a 12-month, 15-department program decisive — one new committee, a handful of part-time roles, no new hierarchy. The program is CEO-led by mandate (see board presentation, decision 2).

Roles

Executive sponsor — CEO (PRIME PRODUCTS)

The program’s authority and its most visible advocate. Concrete responsibilities:

• Chairs the monthly steering committee; owns the program mandate before the board.
• Removes blockers that cross department lines within one week of escalation.
• Personally fronts key communications: kickoff all-hands, quarterly all-company updates, recognition of early adopters.
• Approves stage-gate transitions (with steering) and any scope/budget change beyond steering authority.
• Time commitment: 2–3 hours/week (unblocking, visibility, 1:1 with transformation leads) plus the monthly steering chair (90 min + prep). This commitment is explicit because sponsor fade is the top governance failure mode at this company size.

Steering committee

• Composition: CEO (chair), 2–3 department heads (fixed core: sales director; rotating seat across the other 14 departments), IT lead, TPL transformation lead, Prime Products-side transformation counterpart. Security/data owner attends when data-flow decisions are on the agenda.
• Mandate: approve monthly priorities and the department rollout sequence; allocate budget within the board-approved envelope; accept/reject stage-gate results; own the risk register top risks; resolve escalations; approve any external-facing AI workflow before production.
• Cadence: monthly, 90 minutes, agenda fixed (KPI scorecard → stage/milestone status → decisions required → top risks → next-month priorities). Decisions minuted in the knowledge base; material ones become ADRs in 06-Decisions.

AI transformation lead (TPL) and counterpart (PRIME PRODUCTS)

A deliberate two-lead structure — TPL brings method and build capacity, the counterpart brings organizational reality and is the vehicle for capability transfer.

• TPL transformation lead: runs the program day-to-day; owns plan, WBS, roadmap, and the weekly ops report; directs TPL delivery (platform, knowledge base, Mission Control, training); prepares steering packs.
• PRIME PRODUCTS transformation counterpart (designated by CEO, ~50% time — assumption, to validate): the program’s face inside PRIME PRODUCTS; coordinates champions; owns internal communications execution; arbitrates department scheduling conflicts below steering level; progressively takes over program operation toward the month-12 handover.

IT lead (PRIME PRODUCTS)

Owns platform infrastructure operations: hosting environment, network, backup, monitoring, user provisioning, first-line support (with TPL second-line during year 1). Counterpart to TPL engineering on security and operations. ~30–50% time during build, ~20% steady state.

Department AI champions (15 — one per department)

The transformation’s capillaries: ~10–15% of their time, board-authorized.

• Carry discovery in their department; maintain the department use-case backlog with the transformation leads; test pilots first; train and support colleagues; feed the knowledge base; report adoption blockers weekly.
• Selection criteria (nominated by department head, confirmed by steering): respected by peers, hands-on knowledge of the department’s daily work, demonstrated curiosity about new tools (not necessarily “technical”), available capacity — explicitly not simply the most junior or the department head themselves. Comfortable working in both Greek and English preferred.

Knowledge manager (PRIME PRODUCTS, from Q2)

Curates the Obsidian knowledge base (blueprint): structure and template stewardship, review of champion contributions, coverage tracking against the discovery process inventory, quality gate before content feeds RAG. ~20–30% time.

Process owners

For each process being documented or automated, the department head names a process owner — the person accountable for how that process actually runs. The process owner signs off the documented process, approves any agent operating on it, and owns the human-approval step for its outputs.

Security/data owner

IT lead or compliance/quality head (to be decided at kickoff). Approves every new data flow into the platform (source, sensitivity, access scope); owns the data classification applied to RAG sources; investigates any suspected data incident; reports data-incident status (target: zero) in every steering pack.

Meeting cadence

Meeting	Who	Frequency	Inputs	Outputs
Steering committee	CEO (chair), core members (above)	Monthly, 90 min	Monthly steering pack, KPI scorecard, decision memos, top risks	Minuted decisions, priorities, stage-gate verdicts, escalation resolutions
Program ops sync	Transformation leads, IT lead, workstream leads	Weekly, 45 min	Weekly ops report draft, workstream status, blocker list	Updated plan, blocker assignments, escalation candidates
Champions circle	All 15 champions + transformation counterpart (facilitator)	Bi-weekly, 60 min	Adoption metrics, pilot feedback, department blockers	Shared practices, backlog updates, blockers raised to ops sync
Department working session	Champion, process owner(s), TPL delivery, dept. head as needed	Per pilot, weekly during active pilot	Pilot metrics vs. baseline, user feedback	Pilot iterations, go/no-go recommendation
CEO 1:1	CEO + transformation leads	Weekly, 30 min (within the 2–3 h/week)	Weekly ops report, escalations	Sponsor actions, communication decisions
Board update	CEO (presenter), TPL lead supporting	Quarterly	Quarterly board pack	Board feedback, envelope/stage confirmations

Decision rights matrix

R = recommends, A = approves/accountable, C = consulted, I = informed.

Decision	CEO / sponsor	Steering	Transformation leads	Department heads	IT lead
Program budget envelope, stage gates	A (board for envelope)	R	R	C	C
Monthly priorities & rollout sequence	C	A	R	C	C
Department pilot go-live	I	I	R	A (own dept.)	C
Any external-facing AI workflow to production	C	A	R	C (owning dept.)	C
New data source into the platform	I	I	R	C (data-owning dept.)	A (with security/data owner)
Hardware & platform technical choices	I	C (if budget-relevant)	R	I	A (operability), TPL A (architecture)
Champion nominations	I	A (confirmation)	C	R (nominate)	I
Knowledge-base structure & conventions	I	I	A (with knowledge manager)	C	I
Cross-department resource conflicts	A (if unresolved)	A	R	C	C
Program scope change (charter-level)	R	R	R	C	C
Training plan & wave scheduling	I	C	A	C (dept. timing)	I

Escalation paths

1. Operational blocker (pilot stuck, data access missing, tooling issue): champion or workstream lead → weekly ops sync → resolved or assigned within one week.
2. Cross-department conflict (scheduling, priorities, resource contention): transformation counterpart mediates → unresolved in one week → steering committee → unresolved there → CEO decides.
3. Data/security concern: anyone → security/data owner immediately (not the next meeting) → CEO informed same day for any suspected incident; steering reviews post-mortem.
4. Budget/scope beyond steering authority: steering → CEO → board at next quarterly update, or extraordinary board session if time-critical.
5. People concerns (workload, job-security anxiety): champion or any employee → transformation counterpart or HR → themed in champions circle and steering; the no-headcount-reduction commitment is restated in every such resolution.

Standard: every escalation gets an owner and a response deadline when raised; “parked” is not a state.

Reporting rhythm

Report	Producer	Audience	Frequency	Contents
Weekly ops report	TPL transformation lead	Sponsor, steering members, champions	Weekly (published on Mission Control)	Progress vs. plan by workstream; blockers and owners; decisions needed; next week’s focus; adoption snapshot
Monthly steering pack	Transformation leads	Steering committee	Monthly, 3 working days before steering	KPI scorecard vs. targets (KPI framework); milestone/stage status vs. roadmap; budget consumption vs. envelope; top risks with movement (risk register); decision memos; use-case pipeline status (pipeline)
Quarterly board update	CEO with TPL support	Board	Quarterly	Objective attainment vs. the 8 program objectives; measured business impact vs. baselines; spend vs. envelope; stage-gate results and next-stage request; top-3 risks; outlook
Adoption & platform dashboard	Automated (Mission Control)	Whole company	Continuous	Live KPIs: active users, use cases in production, knowledge-base coverage, availability — the transparency instrument of the program (mission control)

Board update rhythm. Beyond the quarterly cadence: an extraordinary board touchpoint is triggered by any confidential-data incident, a forecast stage overrun >20% of stage budget, or a charter-level scope change. The month-12 board session doubles as program closure review: success criteria verdict (charter § success criteria), handover confirmation, and year-2 portfolio decision.