Three Horizons
Workstream Status
| Workstream | Lead | Branch | RAG | Progress | Horizon | Tier |
|---|
7-S Health Check
Influence Model
Decisions Required
Key Risks
Weekly Notes
Area Overview
People Map
Workload Distribution
Resource Heatmap
Agentic Autonomy — Maturity & Inaction Risk
| Stream | Tier | Knowledge | Workflow | Agent Cov. | Handoff | Maturity | Target | Inaction Risk |
|---|
Agent Coverage
Risk Register
| # | Risk | Impact | Prob. | Owner |
|---|
Risk Heatmap
Pending Decisions
| Decision | Owner | Due | Workstream |
|---|
Decided
| Decision | Owner | Date | Outcome |
|---|
Quarterly Delivery
Pipeline RDA
Budget per Vendor
Budget per Stream
Dettaglio Costi
| Stream | Milestone | Voce | Vendor | Importo | Owner | RDA | Stato | Scadenza | Note |
|---|
Overall Progress
RAG Distribution Over Time
Per-Workstream Trend
Dashboard Guide
This dashboard uses McKinsey frameworks adapted to the TeamSystem ESG program. Below is the explanation of each concept.
Three Horizons
Key question: How much does it hurt if you don't do it by the deadline?
H1 Protect
"If I don't do it, I get fined or blocked."
Regulatory obligations, non-negotiable deadlines.
Examples: NFD, RI, Environment Policy, Forest audit, Supplier CoC.
Horizon: 0-6 months.
"If I don't do it, I get fined or blocked."
Regulatory obligations, non-negotiable deadlines.
Examples: NFD, RI, Environment Policy, Forest audit, Supplier CoC.
Horizon: 0-6 months.
H2 Build
"If I don't do it, I lose an advantage I can't recover."
Infrastructure for the future. Not mandatory today, but if you don't build it now, in 12 months you're behind.
Examples: Taxonomy KPIs, CarbonWallet engine, Workday integration, DMA 2026.
Horizon: 6-12 months.
"If I don't do it, I lose an advantage I can't recover."
Infrastructure for the future. Not mandatory today, but if you don't build it now, in 12 months you're behind.
Examples: Taxonomy KPIs, CarbonWallet engine, Workday integration, DMA 2026.
Horizon: 6-12 months.
H3 Create
"If I don't do it, nothing changes today. But if I do, everything changes tomorrow."
Transformation: changes how the organization works.
Examples: Masterclass for 4500 employees, agentic system, CarbonWallet v2 for 2M+ SMEs.
Horizon: 12+ months for impact.
"If I don't do it, nothing changes today. But if I do, everything changes tomorrow."
Transformation: changes how the organization works.
Examples: Masterclass for 4500 employees, agentic system, CarbonWallet v2 for 2M+ SMEs.
Horizon: 12+ months for impact.
Healthy program: H1 ahead, H2 in progress, H3 started. If you only have H1, you're firefighting. If you only have H3, you're a visionary with no legs.
Tier — Autonomy Level
Key question: Who drives the work, the human or the system?
L1 — Amplifier
The human drives, the system executes in real time. The human thinks and decides, the system does the hands-on work.
Examples: Masterclass (Francesco drives every decision), Taxonomy (Anna decides product by product).
The human drives, the system executes in real time. The human thinks and decides, the system does the hands-on work.
Examples: Masterclass (Francesco drives every decision), Taxonomy (Anna decides product by product).
L2 — Copilot
The human launches, the system does 90%. The human reviews and approves the result.
Examples: Reporting (Lucia launches, agents produce), Stakeholder Engagement (Alice collects, system analyzes).
The human launches, the system does 90%. The human reviews and approves the result.
Examples: Reporting (Lucia launches, agents produce), Stakeholder Engagement (Alice collects, system analyzes).
L3 — Autopilot
The system does everything autonomously. The human validates the result and intervenes only if something goes wrong.
Examples: Automatic priority triage, follow-up scheduling, data pipeline.
Goal: bring as many streams as possible to L3.
The system does everything autonomously. The human validates the result and intervenes only if something goes wrong.
Examples: Automatic priority triage, follow-up scheduling, data pipeline.
Goal: bring as many streams as possible to L3.
The Tier does not indicate quality or importance. An L1 can be the most critical (requires the leader's brain). An L3 is the most mature (the system runs on its own).
RAG Status
Each initiative has a traffic light based on status and deadline:
Green — On track. Status done or in_progress with distant deadline.
Amber — At risk. Planned (not started) or in_progress with close deadline and low progress.
Red — Blocked or critical. Status blocked or imminent deadline with very low progress.
ESG Pillars
Each stream belongs to one or more pillars:
EEnvironment — Climate, emissions, biodiversity, resources. (SpritePinky, Forest, CarbonWallet, TNFD)
SSocial — People, community, inclusion, suppliers. (Open Doors, Inclusion, Suppliers)
GGovernance — Reporting, compliance, platform, processes. (NFD, Taxonomy, Board ESG)
7-S Health Check (McKinsey)
7 dimensions of organizational alignment. Score 1-10 computed from program data:
Strategy — Clarity of direction (milestones with defined "why")
Structure — Stream organization (submodules, dependencies, ownership)
Systems — Automation level (L1/L2/L3, operational agents)
Shared Values — Team alignment (all streams have "why" and contributors)
Skills — Competencies present (coverage of contributor "knows" + agent categories)
Style — Leadership style (approvals, documented decisions)
Staff — Workload distribution (single points of failure)
Structure — Stream organization (submodules, dependencies, ownership)
Systems — Automation level (L1/L2/L3, operational agents)
Shared Values — Team alignment (all streams have "why" and contributors)
Skills — Competencies present (coverage of contributor "knows" + agent categories)
Style — Leadership style (approvals, documented decisions)
Staff — Workload distribution (single points of failure)
Influence Model (McKinsey)
4 levers for driving organizational change:
Role Modeling — Leaders do what they ask others to do. (Does Francesco use agents? Does the ESG team use Ground Control?)
Understanding — People understand the why. (Does every milestone have a "why"? Does the masterclass explain ESG?)
Reinforcement — Structures and processes support the change. (Impact platform, agents, automated pipelines)
Talent — The required competencies are present. (Knowledge map, documented handoffs, training)
Understanding — People understand the why. (Does every milestone have a "why"? Does the masterclass explain ESG?)
Reinforcement — Structures and processes support the change. (Impact platform, agents, automated pipelines)
Talent — The required competencies are present. (Knowledge map, documented handoffs, training)
How progress is calculated
Progress for each milestone comes from the average of its specs:
done = 100% | in_progress without explicit % = 30% | planned = 0% | blocked = 0%
If a spec has
Workstream progress is the average of its milestones. Horizon progress (H1/H2/H3) is the average of initiatives in that horizon.
The global number is an average of all initiatives and hides variance. Always look at the 3 horizons separately.
If a spec has
progress: 50% or progress: 3/8, that exact value is used.Workstream progress is the average of its milestones. Horizon progress (H1/H2/H3) is the average of initiatives in that horizon.
The global number is an average of all initiatives and hides variance. Always look at the 3 horizons separately.
AgenticOS Vision
Ground Control is not a project tracker — it is the operating system of the ESG team.
YAML
= Process definitions
= Process definitions
Agents
= Worker processes
= Worker processes
Dashboard
= Control panel
= Control panel
Effort formula: effort_umano = L1*1.0 + L2*0.3 + L3*0.1
Only human effort counts for capacity planning. Agent effort has near-zero marginal cost post-setup.
Autonomy Maturity — 4 Pillars:
Knowledge Infrastructure (30%) — Specs, docs, KB available for agents
Workflow Standardization (25%) — Repeatable process, defined steps
Agent Coverage (25%) — % milestones covered by agents
Handoff Friction (20%) — Smoothness of human-agent transitions
Workflow Standardization (25%) — Repeatable process, defined steps
Agent Coverage (25%) — % milestones covered by agents
Handoff Friction (20%) — Smoothness of human-agent transitions
GC Advisor Agent
A single agent with two lenses — strategic (what, why) and operational (who, how much, how).
Strategic Lens:
7-S • Influence Model • Three Horizons • MECE • Critical Path
ESG Frameworks:
CSRD Readiness • Taxonomy Alignment • Stakeholder Coverage • Knowledge Risk • Regulatory Timeline
7-S • Influence Model • Three Horizons • MECE • Critical Path
ESG Frameworks:
CSRD Readiness • Taxonomy Alignment • Stakeholder Coverage • Knowledge Risk • Regulatory Timeline
Operational Lens:
7 Agent Categories • Resource Heatmap • Agent Coverage Audit • Transition Planning
Hybrid team model:
~10 people + ~20 agents across 12 streams. Effort:
7 Agent Categories • Resource Heatmap • Agent Coverage Audit • Transition Planning
Hybrid team model:
~10 people + ~20 agents across 12 streams. Effort:
L1*1.0 + L2*0.3 + L3*0.1
Agent Coverage Model
7 agent categories define what kind of automation each stream needs:
| Category | Function | When needed |
|---|---|---|
| Orchestration | Coordinates workflow across agents and people | 3+ interdependent milestones |
| Compliance | Verifies regulatory/standard adherence | Regulatory constraints (reporting, taxonomy) |
| Quality | Checks data, numbers, consistency | Measurable output |
| Narrative | Writes, reformulates, translates for audience | Communicative output (RI, masterclass) |
| Tooling | Builds/maintains tools | Platform streams (ts-impact-platform) |
| Engagement | Manages stakeholder interactions | Relational component |
| Tracking | Monitors, collects, catalogues data | BAU repetitive activities (operations) |
Agent suggestion logic: If a milestone is repetitive + structured + data available = high agentizability (L3 candidate). If creative + decisional = stays L1.
Transition path: L1 (human guides) → L2 (human launches, agent does 90%) → L3 (agent does everything, human validates).
Transition path: L1 (human guides) → L2 (human launches, agent does 90%) → L3 (agent does everything, human validates).