AI strategy and use cases
Translate priorities into a limited portfolio with owners, success measures and stop conditions.
Governance and responsible use
Define authorised tools, data rules, review responsibilities, incident handling and periodic reassessment.
Adoption and capability building
Engage users in design, training and feedback so that validated practices become part of real work.
How Consultor and Monitor contribute
Marouane Harmach leads the advisory approach through Consultor. Monitor contributes when media intelligence, strategic monitoring or reputation are part of the decision.
From strategy to deployment
Continue with the AI maturity and prioritization audit, the ROI and scaling guide, and use cases by business function. They complement strategy, governance, adoption, and training without duplicating search intent.
Train your leadership team
See the in-company executive AI training program in Morocco, including formats, deliverables, and scoping with Consultor.
AI strategy for Moroccan organisations
An AI strategy is a set of management choices. It explains why the organisation should act, which decisions and activities deserve attention, what information can be used, who remains accountable and how evidence will be reviewed. A credible roadmap avoids treating every idea as a project. It maintains a limited portfolio and states why each use case is being explored.
In Morocco, the operating context includes the organisation’s digital maturity, languages of work, sector obligations, availability of reliable documentation and ability to support users. The Moroccan CNDP stated in March 2025 that AI processing involving personal data is governed by Law 09-08. This makes data purpose, access, transparency and human accountability early design questions rather than a final compliance check.
Selecting AI use cases
Each opportunity should describe the user, current task, expected change, required data, acceptable error, human review and dependency on other systems. Use cases can then be compared across value, feasibility, risk and adoption. A short scale with a written justification is more useful than a precise-looking score whose assumptions remain hidden.
A first experiment should answer a decision. It may test whether people can use an assistant appropriately, whether the available documents support reliable retrieval, or whether the correction effort removes the expected gain. The output is a documented choice to continue, change or stop.
Governance for generative AI
Governance establishes authorised tools, information rules, owners, validators, incident handling and review dates. The level of control should match the consequence of an error. Drafting assistance and decisions affecting people do not require the same review. A use register helps leadership see where AI is used, for what purpose and under whose responsibility.
Adoption and executive support
Adoption is designed with the people who perform the work. Training should use realistic tasks, show limitations, explain escalation and give teams a way to report problems. Executives need a shared language for value, risk and evidence so that investment decisions are consistent across departments.
Typical deliverables
| Workstream | Deliverable | Decision supported |
|---|---|---|
| Opportunity mapping | Prioritised use-case portfolio | Where to start |
| Governance | Use register and decision rules | What can be authorised |
| Experiment | Protocol, baseline and review | Continue, adjust or stop |
| Adoption | Training and support plan | How to sustain appropriate use |
Public evidence
Interviews, conference records and source links are organised in the AI media evidence library. They document public activity and the roles under which publishers presented Marouane Harmach. Consultor is the vehicle for advisory assignments; Monitor contributes when information monitoring or reputation is part of the problem.