Ethics focused on people
Technology at the service of digital rights and people’s well-being.
KarminaAI StudioAgents · Services · Training
Public administration · AI Master Plan 2026–2028
Govern change before scaling it.
An artificial intelligence strategy built with the municipal workforce to move from dispersed uses to a shared institutional capacity, safe and oriented to public value.
A social contract and a strategic compass for an augmented city.
Approved by the Municipal Assembly04·2026
InstitutionTerrassa City Council
ProjectAI Master Plan
Period2026–2028
AbastStrategy · Governance · Training
El challenge
AI was already there. What was missing was a shared way to decide.
The interest in learning was high, but the uses had started to grow individually: non-corporate tools, different criteria and few common references to data, purposes or supervision.
The challenge was not to incorporate more technology. It was necessary to convert a transformation that was already underway into an institutional capacity: with a vision, an architecture of responsibilities and a sure way to move from an idea to a service.
Before scaling anything up, the City needed to know where it was, what it wanted to protect, what opportunities had public value, and how it would prepare people who should use and monitor future systems.
“The question was not whether AI would enter the City Council, but how to govern its entrance and put it at the service of people.”
Institutional change
A policy built from within the organisation.
The participation allowed the needs, frictions and risks perceived by the staff to form part of the strategy. This organizational legitimacy is as important as the technological definition.
A CAS IS NOT COPY
We can understand what is transferable in your context.
In a free video call we will separate the useful idea from the one that depended on the client, the data, the team or the specific moment.
Video call · free · no obligationBook a free video call ↗A short conversation to understand your starting point and bring clarity.
A process in five movements
From an internal assessment to an approved policy.
Diagnosis, management alignment, participation and strategic definition were not decorative phases. Each movement reduced uncertainty and prepared the next.
- 00
Diagnosi
Radiography of the uses, capacities, needs and risks to build the Plan on the reality of the organization.
- 01
Alignment
A day with commands and management profiles to share language, opportunities, limits and responsibilities.
- 02
Participation
Survey, webinar, work groups and Ideaton to turn everyday frictions into concrete needs and proposals.
- 03
Strategy
Principles, governance, ethical and regulatory framework, literacy, monitoring and roadmap for the period 2026–2028.
- 04
Approval
The municipal plenary approves the Plan in April of 2026 and turns it into a shared institutional policy.
Responsible for the projectLead the project
Ignasi Llorente
Partner Director UtopiqDirecting a municipal AI Plan is not ordering an inventory of tools. It is to hold a conversation between strategy, technology, data, rights, organizational culture and public service until these dimensions become applicable decisions.
In this project, the leadership focused on three issues: honestly understanding the starting point, involving the people who know the day to day of the City Council and leaving an architecture that could continue to work after approval.
Listen before prescribingTranslate complexity into criteriaAbility to leave within the organization
Discover Utopiq ↗The architecture created
Six pillars. One guiding principle: public value.
Efficiency of services
Less repetitive load without lowering the quality or guarantees of the public service.
Governance of data
Quality, interoperability, protection, traceability and proper use of municipal knowledge.
Internal Capabilities
Teams ready to use, monitor, evaluate and govern AI systems.
Citizen Confidence
Transparency, explainability and accountability as deployment conditions.
Comprehensive sustainability
Public benefit provided to the economic, energy and environmental resources that the solution consumes.
PUBLIC VALUE
AI only makes sense when it improves a service, strengthens institutional capacity, or generates real benefit for the community.
The engine of governance
Shared responsibility. Clearly accountable decisions.
The Plan connects three levels so that the strategy is not detached from the operation and so that the day-to-day needs can reach the decision-making bodies.
CIAT
It defines priorities, validates projects with greater risk and guarantees ethical, legal, organizational and political alignment.
Municipal Office of Data
It coordinates pilots, keeps the record, works on data governance and ensures the application of the criteria.
AI References
They detect opportunities, collect needs and connect the general strategy with the reality of each service.
From use case to public service
An idea does not go straight into production.
Each proposal goes through a life cycle that allows to stop, correct or withdraw the solution before it reaches the operation.
- 1
Detect
An area identifies an operational friction or a specific citizen need.
- 2
Screen
The proposal is reviewed according to risk, data protection and internal criteria.
- 3
Design
Purpose, data, supervision, guarantees and indicators are defined.
- 4
Pilot
The solution is tested in a limited, monitored and reversible environment.
- 5
Scaling
Only what works is integrated, recorded and reviewed continuously.
YOUR START POINT
Tell us what you want to transform.
A summary of the challenge, the current process and the expected result is enough to begin to value a route of its own.
The project in figures
Participation also builds infrastructure.
These data explain the scope of the process, not a technological impact that has not yet occurred. The identified 25 proposals are pending prioritization and are not presented as implanted pilots.
the internal survey on knowledge, use and perception of AI
the day of training and alignment for commands
the Ideaton Preparatory Webinar
at the Ideaton session
Identified and prioritised; not yet implanted pilots
until the first draft of the Master Plan is available
to connect strategy, governance, ethics, regulation, training and monitoring
to consolidate foundations, deploy pilots and scale only what shows value
Transformation of Work
Reduce the burden of repetition. Give more weight to judgement.
You can lose weight
Routine processing
Manual classification
Standard reports and summaries
You have to gain weight
Critical thinking
Empathy and support
Policy design and ethical oversight
Training according to role and responsibility
A common foundation for the entire workforce; an intermediate level for managers and technical profiles; and advanced training for AI referents, technology and legal fields.
View training on AI ↗The Red Lines
Oversight increases with risk.
Internal support
Organization, assisted writing or documentary classification with controlled sources and human review.
Interaction with people
In a citizen care assistant, the person must know that they interact with a system of AI.
Essential rights and services
Selection, grants or other sensitive areas require strengthened evaluation, authorization and supervision.
Outside the frame
Social score, manipulation of vulnerable people and certain unauthorized biometric uses.
What comes next
Foundations. Pilots. Scalability.
- 2026
Foundations and Governance
Bodies, municipal register, first protocols and start of the literacy program.
- 2026–27
Pilots and Integration
Controlled tests, monitoring and adjustments based on evidence.
- 2027–28
Scalability
Transversal expansion only of solutions that have demonstrated value and guarantees.
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 companyWhat does Terrassa teach us
An institution needs to know what it wants to transform, and under which rules, before choosing a tool.
The Plan leaves a navigation chart: an honest diagnosis, shared priorities, defined responsibilities, an initial portfolio of proposals, level training and a mechanism to test without losing control.
This is the most important result of the project: Terrassa can continue to advance without confusing speed with improvisation.
Follow the thread
From this case study to other starting points.
Expertise · 01
AI for public administration
Strategy, governance, knowledge, pilots and compliance adapted to public service.
Explore ↗Services · 02LLM Strategy
Decide where the AI can add value and turn that decision into an applicable roadmap.
Explore ↗Knowledge · 03Knowledge Foundation
Order sources, permissions, vocabulary, risks and accountability before connecting to any agent.
Explore ↗Learning · 04AI training
Itineraries in partner adapted to the role, the level of departure and the real responsibilities.
Explore ↗Sources and reading criteria
A documented case with verifiable process data.
Published figures describe participation, scope and capacity created. The identified proposals have not been converted into implementation results.
News of the approval of the Plan ↗Strategic Innovation Service Executive Summary ↗Let’s talk
What should your organization be better able to decide?
Tell us about your situation, even if the challenge is not fully defined. A member of our team will respond to understand the process and discuss a first step.
hello@karmina.ai