Management and business
Vision, objectives, sponsor, business plan, investment criteria and indicators to decide what goes on and what stops.
KarminaAI StudioAgents · Services · Training
STRATEGIC CONSULTATION · PLA AI DIRECTOR
AI is already entering your company. The plan decides how.
A process of direction, governance and adoption to align the entire organization: from business opportunity and internal knowledge to tools, risks, training and execution.
PLA
DIRECTOR
AIStrategyKnowledgePeopleGovernanceTechnologyRisksExecution
DIRECTOR
AIStrategyKnowledgePeopleGovernanceTechnologyRisksExecution
The usual horizon6–12 Months
With a limited first phase of 3–Z6 months when scope allows.
OrganizationFrom 50–100 people
AbastTransversal · not only technology
ResultDecisions, deliverables and internal capacity
ConditionReal commitment of management and teams
NOT A DOCUMENT TO ARCHIVE
Everything your company needs to think through, coordinate, align and present so AI is used well.
When each person tries a tool, the speed seems to progress. But without shared criteria appear duplicate licenses, fragmented knowledge, data out of control and use cases that nobody knows how to prioritize.
The Master Plan turns this dispersed movement into a business capacity: a common vision, justified technological decisions, clear responsibilities, trained teams and a portfolio of initiatives that can go from idea to execution.
THE COST OF NOT COORDINATING
Seven tools. Four sets of criteria. No owner.
Personal LicenseFree AccountSensitive data?Pilot without metricPrompt in a chatNo one reviewsWHO
DECIDES?
DECIDES?
This is what the plan calls for. Not by banning experimentation, but by turning it into shared learning.
Management stops seeing a collection of evidence. Technology ceases to receive isolated demands. Teams know what they can do, with what, with what data and when they should ask for help.
BEFORE YOU CLOSE ABAST
You do not need to arrive with a defined solution.
We can review the challenge, starting point and priorities in a free video call. Putting order before writing the scope usually saves time and budget.
Video call · free · no obligationBook a free video call ↗A short conversation to understand your starting point and bring clarity.
A TRANSVERSAL LOOK
The plan connects conversations your company usually has separately.
Business, technology, people, legal, data, communication and operations must be able to look at the same map, even if they do not make the same decisions.
Processes and use cases
Audit of frictions, portfolio of opportunities, prioritization and tokens so that an idea is no longer just a promise.
Knowledge
Sources with authority, responsibility, permissions, validity and an architecture so that knowledge can be consulted and maintained.
Governance
Committees, roles, registration, Human Gates and approval circuits so that the responsibility is not diluted.
Risks and Compliance
Map and matrix of risks, impacts, transparency, data protection, incidents and controls proportional to each use.
Technology and suppliers
Objective architecture, licensing policy, purchasing criteria and selection of tools without dependence on a specific brand.
People and adoption
Literacy, role training, Ideathon, communication, internal referents, support and follow-up after launch.
A Master Plan doesn’t just say what we’re going to do with the AI. Make it clear who can propose it, who validates it, with what knowledge, what risk we assume and how we will know if it works.
IT'S FOR YOU IF...
You want to build capability, not just buy a document.
- Management is willing to sponsor and make decisions.
- The areas will share processes, frictions, uses and limits.
- There will be time to form, test, review and discard.
- Accept that not all ideas deserve a tool or an agent.
- You want to measure business, quality, risk, experience and ability.
DELIVERY MULTS. ONE SYSTEM.
The plan provides the framework. Underneath it are the components that make action possible.
We do not release an entire generic list by inertia. We define which pieces the organization needs, in what order and who will have to maintain them.
To
Management, business and portfolio
To turn ambition into decisions that can be defended in front of direction.
- ↗Audit of the starting point and map of maturity
- ↗Vision, strategic principles and objectives of the Plan
- ↗Business Plan from AI and return criteria
- ↗Map of processes, frictions and opportunities
- ↗Priority portfolio of use cases
- ↗Roadmap, budget and dependencies
- ↗Table of command and continuity criteria
B
Governance, ethics and risk
Because every use has owner, limits, control and a response if something fails.
- ↗Ethical Code of Artificial Intelligence
- ↗Internal Manifest for Responsible Use
- ↗External manifesto and public commitments
- ↗Governance model, RACI committee and matrix
- ↗Registration of systems, uses and suppliers
- ↗Risk plan and risk matrix by use case
- ↗Human Gates and autonomy levels
- ↗Incident Protocol, Review and Withdrawal
- ↗AI Act, GDPR criteria, transparency and documentation
C
Knowledge and technology
To choose an architecture that works with real knowledge and does not leave it scattered.
- ↗Audit of knowledge and corporate sources
- ↗Plan of Knowledge Foundation
- ↗Architecture of the Karmina Company Brain
- ↗Map of owners, permissions, versions and expiration
- ↗Data and knowledge policy for AI systems
- ↗Targeted Technology Architecture
- ↗Licences, Tools and Exceptions Policy
- ↗Shortlist, testing and selection of technology
- ↗Blueprints of agents, services and priority integrations
D
Adoption, communication and monitoring
Because the Plan changes behaviors, activates teams and continues to live after presenting it.
- ↗Literacy and role-based training plan
- ↗Problem-oriented interdepartmental ideathon
- ↗Internal communication plan on AI
- ↗External communication plan with AI
- ↗Social media plan for the team
- ↗Daily guidelines and criteria
- ↗Referral network, community of practice and support
- ↗Pilot files, onboarding and adoption
- ↗Cadence of follow-up, revisions and updating of the Plan
CHOOSE TECHNOLOGY IS ALSO STRATEGY
A company can have several tools. A department should not have several sources of chaos.
What is not negotiated every time
Identity, security, data, purchases, knowledge, permissions, registration, support and risk criteria.
The right tool for the problem
Models, suites, automation or specialized tools are compared according to need, integration, total cost and internal capacity.
Diversity with a written reason
An exception may make sense. Must have owner, permit, review, expiration date and exit plan.
1 department→
an approved catalog, a licensing policy, and a clear door to ask for exceptions.
Explore the Toolkit ↗KNOWLEDGE · THE FIL THAT CONNECTS IT
If AI does not know what your company knows, it will work with fragments.
The Plan identifies the sources that have authority, who holds them, who can see them, when they expire, and what gaps prevent them from moving forward. This layer of Knowledge Foundation prepares the architecture of the Karmina Company Brain.
Knowledge is not an annex to the Plan. It is the infrastructure that allows people, tools and agents to work with the same reality.
Karmina, the agency behind the Studio
Why Karmina?
Behind Karmina AI Studio there is an agency with track record, team, scale and applied experience in real organizations.
Karmina · AI Studio · Criterio · Dunder Mifflin · Utopiq · Smart Team
+€119mIn advertising investment for clientsAds Platforms
Partners
MetaGoogle Ads
TikTok
Pinterest
BarcelonaMadrid
Buenos AiresOur offices · Meet the team ↗
+50 KCommunity
B CorpCertified companyIF YOU ALREADY HAVE MATERIAL
Do you already have a brief? Send us.
It can be a finished document, a few notes or an email with context. We will read it before proposing a conversation or a first step.
IDEATHON · FROM PROBLEM TO DECISION
We bring the right teams together to solve real problems, rather than gather the whole company to imagine the future.
The idea comes after a previous base. It can convene a selection of departments, managers and operational profiles. Thus, ideas are born with knowledge of the process and can proceed to execution.
We come in context.
Interviews, audits, data and real frictions. Let’s not start with a rain of blank ideas.
We work problems
Teams from different areas reshape challenges, contrast value, risk and feasibility, and propose concrete answers.
Let's decide
Each proposal comes with owner, next step and fits: training, tool, agent, service, pilot or discard.
TAKER · SHAPER · MAKER
Maturity is assessed process by process, rather than attached to a company as a label.
The map helps to not build more technology than necessary and to decide what should come out of the Ideathon.
Adopt
A team can solve the problem with existing tools, approved, well configured and with clear criteria of use.
Possible exit · Training + ToolkitAdapt
It is necessary to connect a tool with the process, sources, rules and supervision of the organization itself.
Possible exit · Service + integrationBuild Capacity
The case justifies agents, automations or own architecture because there is sufficient knowledge, volume, risk and value.
Possible exit · Agent + governed systemA 6–12 MONTH SERVICE
The plan is presented. The capability is built.
In 3–6 months, a clearly scoped project can be completed and key decisions established. When a project spans departments, includes activation and aims to demonstrate adoption, the typical timeline extends to 6–12 months.
- 000–1 month
Commitment and Perimeter
Sponsor, motor team, objectives, access to information and success criteria.
- 011–3 Months
Diagnosis and Listening
Audits, interviews, surveys, inventory of uses and reading of risks, knowledge and skills.
- 022–4 Months
Alignment and Ideathon
Common training, interdepartmental work and prioritization of problems that deserve an answer.
- 033–6 Months
Plan Design
Governance, principles, technology, portfolio, training, communication, budget and roadmap.
- 046–12 Months
Activation and tracking
Pilots, selection of tools, agents or services, onboarding, metrics, review and adjustment of the Plan.
SCHOOL EXPERIENCE
We have worked on master plans for organisations of 5, 100, 1,000 and 10,000 people.
Where does the full service fit best?
From around 50–100 people, when there are enough departments, decisions, tools and dependencies for coordination to deliver clear value.
A RED LINE
If the only aim is to cut costs and let people go, we are not the right team.
Productivity matters. They also matter the margin, speed and efficiency. But a Responsible Management Plan decides what we will do with the capacity that is released: improve quality, better care, reduce risk, innovate, grow or return time to the teams.
The purpose is not a phrase at the beginning of the document. It is seen in the budget, in the use cases that are prioritized, in the indicators and in the decisions that the company refuses to automate.
A CASE THAT MAKES THE METHOD TANGIBLE
Terrassa: listening, training, prioritising and leaving a structure that can keep evolving.
Terrassa City Council’s AI Master Plan 2026–2028 shows how a complex organisation can move from scattered uses to a shared policy. Context is public; work logic is applicable to any organization that needs to align people, processes, knowledge, technology, and governance.
Read the full case ↗
A master plan doesn’t have to tell everyone what tool to open. It must make clear what we want to transform, what decisions we will make in common and how we will turn real problems into initiatives with responsibility, risk and follow-up.
REFERENCE BRAND
Governance and literacy serve a real purpose.
The design of the Plan translates frameworks such as the European AI Act and the NIST’s AI Risk Management Framework into decisions, roles, training, records and controls that the organization can apply.
AI Act · Article 4 on ↗ literacyNIST AI RMF · Government, Map, Measure, Manage ↗Let’s talk about the master plan.
Is your company ready to commit to it?
Tell us the size of the organization, the current uses and what a Master Plan has made you think. We will help you to assess if this is the right service or if it is convenient to start by training or a more limited phase.
hello@karmina.aiFAQS · STRATEGY AND KNOWLEDGE
Frequently asked questions about ai master plan.
Five essential answers. The Complete Guide in Development 20.01What is an Artificial Intelligence Master Plan?
+
It is the framework that orders the adoption of the AI for a certain period. It includes vision, principles, diagnosis, use case portfolio, knowledge and technology architecture, training, governance, budget, roadmap and indicators.
02When does an organization need a master plan?
+
It makes sense when several areas already experiment with AI, there are shared budgets or risks or management wants to move from testing to organizational capability. Also before a large investment in platforms.
03Which blocks should you order before defining the roadmap?
+
It is necessary to organize objectives, principles, current situation, contracted tools, critical knowledge, skills, risks and initiatives. Common definitions must also be resolved: what counts as a use case, pilot or system in production.
04How do you incorporate budget and investment scenarios?
+
Investment includes diagnosis, governance, knowledge, platforms, integrations, pilots, training and maintenance. It should be expressed by stages and phases, with conditions to continue, stop or rethink.
05How much should be reviewed and updated?
+
It should be reviewed periodically and when regulations, suppliers, priorities or architecture change. The document must be alive, but the changes must be traced and approved.