Efigence: AI in the everyday work of a design team

In one sentence

I am bringing AI tools into the everyday work of a design team at an agency serving banks, together with rules on what can be handed to a model and what stays with a person.

The problem

AI usually enters design teams chaotically: a few people use the tools their own way, the rest watch, and nobody knows what data went into the model or who is accountable for the result. At an agency working for banks, that last point is a real risk.

What I did

• AI Sagres: I took part in launching this initiative, the group that leads AI projects at Efigence. The idea and the name were mine and come from my talk at AI Lessons Unlearned: designers, business analysts and technology people at one table, the way Prince Henry the Navigator brought people from very different worlds together at Sagres.
• Skills Factory: a library of tools built on an adapter pattern that carries the design system from Figma into code. The team gets a concrete tool for its own work instead of a generic "use AI".
• One line of work with business analysis: we joined forces with the business analysis team's AI initiative, so we would not build two separate islands.

The team and the state of adoption

About 10 people in 2 teams. Each team has its own lead, and one of the designers also specialises in user research. The team delivers projects for banking: Credit Agricole, Alior, Coop Bank.

Roughly 4–5 people out of 10 use AI tools in their daily work, at very different levels. At one end, one senior built a two-way pipeline between Claude Design and Figma and uses generative AI for illustrations, working on this with his team lead. The others mainly use Figma Make or Figma's agents to generate quick early concepts.

How I bring the rest of the team along

The main barrier is a lack of skills and knowledge, so adoption starts from a specific task of a specific person. Adoption is gradual and voluntary.

In September 2026 a designer asked for help checking whether her design met all the requirements in the analysts' specification. We showed her how to do it with AI, and she saw on her own task that it made her work easier.

A second concern came up along the way. She has a fine arts background and believes AI produces uncreative work because it relies on typical patterns. That is an accurate observation and we agree with it. So we showed her a way of working in which creativity stays with her, while AI shortens the path from concept to the first version of a design.

Client data

Efigence has an official document: a policy on the use of AI in its projects. The team follows it to protect clients' assets and knowledge, as well as our own.

What this project proves

That I can take a design team through a change of tools without losing control over quality and accountability, in an industry where client data is a regulatory risk.