Internal memorandum · Confidential
- To
- Board of Directors / Executive Committee
- From
- Chief Executive Officer
- Date
-
- Subject
- Proposal to install a Native AI Department
1. The situation
Artificial intelligence is already inside the company, even if we never decided it: people using assistants on their own, vendors offering "AI solutions" and pilots that never reach production. What we don't have is a place for it: an owner, rules, a trained team and systems connected to the operation.
Most companies in our industry are at the same point. Over the next two years, the difference won't come from who buys more tools, but from who turns AI into a company capability before everyone else.
2. The risk of not deciding
- Money without measurable return: isolated tools and pilots that nobody can evaluate.
- Data risk: customer and company information uploaded to tools with no policy or control.
- Dependency: systems that only an outside vendor understands.
- Falling behind: the team's learning curve starts when we start, not before.
3. The options we evaluated
| Option | What it solves | What it leaves open |
|---|
| Strategy consulting firm | Diagnosis, benchmark and roadmap. | Nobody builds. The roadmap depends on someone else executing it. |
| Software firm or integrator | Working systems. | No strategy or governance. The team doesn't learn and the company depends on the vendor for every change. |
| Courses and training | People who understand what AI is. | Nothing connected to operations. Enthusiasm fades without a first result. |
| Hiring an internal team | In-house capability, in the long run. | It takes time to build, and nobody knows whom to hire first, how to lead them or what to build. |
Each option solves one part. AI capability needs all four pieces at once: strategy, talent, governance and infrastructure.
4. The proposal
Install a Native AI Department: an internal capability with strategy, team, rules and systems, built in stages and with measurable results from the first one. When it's done, the capability belongs to the company.
| Pillar | What gets installed |
|---|
| Strategy | Maturity and operations assessment: where the money is going today. Portfolio of use cases prioritized by return and by ease. 12-month roadmap with each case's metric and its baseline. |
| Talent | Role map: who owns AI, who operates it and who approves in each area. Training by level: leadership, middle management and operations, built on the company's real cases. Usage guides for each area's day-to-day work with the tools. |
| Governance | AI usage policy: what data can be used, in which tools and with which permissions. Decision matrix: what AI resolves on its own and what requires human approval. AI committee with an agenda, indicators and traceability for every system in production. |
| Infrastructure | Data layer connected to our ERP, CRM and critical spreadsheets. AI agents and custom systems in production, starting with the highest-return case. Command Center with the department's indicators and monitoring of every system. |
5. How it would be done
- 11D Scan. A measured starting point: maturity, operations, prioritized cases and roadmap.
- Ignition. The first case in production with a written metric, and the team learning on something real.
- Command Center. The department running: governance in force, several cases in production and their indicators in plain sight.
- Orbit. Handover: our team operates and decides; 11D supports, measures and builds what comes next.
Each stage has a metric agreed in writing before it starts. If a stage doesn't deliver, the next one isn't approved.
6. How we will evaluate it
- Measured baseline of the problems to be tackled, before investing in systems.
- Savings or revenue attributable to each case in production, against that baseline.
- People trained and operating AI, by area.
- Usage policy in force and zero data incidents.
7. Decision requested
Approve the first stage —the maturity and operations assessment (11D Scan)— to obtain the baseline, the portfolio of prioritized cases and the investment proposal for the following stages. Its cost is fully credited toward the first implementation.