Some projects are born from a sudden idea. And some are born from accumulation. Karmina AI Studio clearly belongs to the second group.
Agència Karmina has just unveiled Karmina AI Studio, a dedicated space for turning all the AI knowledge the team has built up over the past few years into agents, services, training programs, and tools. I've had the chance to accompany this phase from the inside, as the team's trainer and consultant, and I want to explain why this launch feels especially significant to me.
None of this happened overnight. Behind it are more than 100 training sessions on LLMs and applied AI, delivered together with Utopiq, aimed at both internal teams and clients across very different sectors. Behind it are also months of real-world experimentation, not lab work: projects carried out with Solorelatio, AENA, and the Ajuntament de Terrassa. Private companies, social organizations, public institutions, and multinationals, all going through the same process of learning to integrate AI without losing control over what actually matters.
At some point along that road, the question stopped being how much we know and became what we do with what we know. Karmina AI Studio is the answer to that question.
The most interesting thing about Karmina AI Studio is what it has decided not to be. It's not a technology catalog. It's not a collection of generic tools labeled "AI" to look current. Above all, it's a way of connecting artificial intelligence with each organization's knowledge, processes, and real needs, while always keeping judgment and the decisions that matter in people's hands.
That last part isn't a minor detail. It's probably the most important decision behind the project. At a time when many companies automate first and think later, Karmina has chosen the opposite path: understand first, automate with judgment after.
Before bringing this philosophy to clients, Karmina applied it to itself first. The first changes were rolled out directly within the management and administration team, together with Martina Díaz-Corbalán and Ariadna D. Miranda. It wasn't a theoretical exercise or a polished demo. It meant redesigning real day-to-day work processes, using voice as the main interaction channel, and relying on tools like Codex and Claude to build an ecosystem that responds to the team's actual needs, not a trendy feature list.
Watching that process up close, as a consultant and trainer, has confirmed something I firmly believe: AI training only makes sense when it translates into real decisions and processes within an organization. Not in slide decks. In the day-to-day.
In 2026, Karmina has doubled its revenue. That's a figure worth mentioning. But to me, it isn't the most relevant one. What really matters is that, instead of simply repeating what already worked, the team decided to build something new from what it had learned. Choosing your own direction instead of following someone else's is almost always the hardest, and most important, decision a growing organization can make.
Karmina AI Studio is only the beginning of this new stage. In the coming months, the team will keep sharing its expansion plans for the years ahead. And I'll keep accompanying that journey with the same conviction we started with: artificial intelligence doesn't replace people's judgment. It amplifies it, when implemented well.