Process. How a project works with us.
Before you start, you should know what you get, how long it takes and who is accountable. Here are the answers.
Fast track. One question, a clear answer.
When you need to decide whether to go on, stop or change direction, and you don't want to wait a quarter. A fixed scope and a small senior team, two days to a month. After it you can stop, or go on to a program.
Direction workshop
What you get: prioritized goals or first concepts, and a summary with recommendations. When: when stakeholders don't agree on the direction of a new product or service.
Synthetic user research
What you get: your audience's needs and behaviors, from AI personas built on real data, ranked by impact and certainty. When: when you need to understand an audience fast, before spending on fieldwork. It shortens research with people; it doesn't replace it.
Concept validation
What you get: where users get stuck in the prototype, and what to fix. One to two weeks with a synthetic test, two to four with users. When: before you commit a build budget.
Digital product diagnostic
What you get: a prioritized list of fixes, against business goals and user needs. When: when an existing app or site isn't delivering what you expected.
Feedback and inquiry analysis
What you get: the topics that keep coming up in reviews, inquiries and surveys. The AI reads everything; people interpret. When: when there's a lot of feedback and nobody has time to read it.
Market and competitor scan
What you get: trends, a comparison with competitors, and the places where you can be different. When: before a strategic or product decision.
Program. A team on a product, over time.
A product that is a core channel doesn't end at launch. Someone has to hold the strategy, the design and the development, and improve it based on what happens in the field. Usually three months or more.
Product strategy and concept
Vision, principles, prioritized features and a live prototype. This is how a program usually starts.
Spec and detailed design
One spec that keeps being updated: requirements, user stories, edge cases and multilingual microcopy. Alongside it, every screen in every state, ready for development.
Design system across all assets
One design language for apps, sites and forms, and a rollout plan in phases.
Embedded product team
Product managers, designers and engineers working inside your teams, with early feedback from development, legal, regulation and data.
AI adoption program
Leadership first, then unit managers, then product teams: personal AI workflows, an agreed way of working with IT and information security, and AI inside research, spec, design, development and QA.
AI in the products we design. With a person who decides.
A digital assistant that decides a sensitive request on its own creates a problem. So we define every touchpoint up front: what the AI does, what a person decides, and how the handoff works without the user feeling a seam.
What the AI does
- Summarizes a long file into a short snapshot
- Sends a request to the right person the first time
- Pre-fills what is already known, without asking again
- Sends a reminder before someone gets stuck
What stays with a person
- Decisions with regulatory weight
- Sensitive situations
- Cases that fit no template
What stays with you.
At the end of a project, a lot of knowledge sits in people's heads. We document it: decisions, constraints and history, in a format both people and AI agents can work from. Your team carries on alone, and comes back to us only when needed.
Three stages. And what AI changes in each.
01 Mapping and strategy
We learn the organization, its history and its users, and decide what we will measure. What AI changes: it summarizes interviews, documents and data quickly, and builds synthetic personas that narrow the questions. What stays with people: the interviews themselves, deciding what matters, and aligning stakeholders.
02 Spec and validation
We specify and design, build a working prototype and test it in the field before a wide rollout. What AI changes: variations and edge cases fast, a prototype within days, and a first test against synthetic personas. What stays with people: the direction, the taste, and testing with real people.
03 Implementation and support
We support development, measure, and document what we learned. What AI changes: automated tests and documentation, and agents that follow your methodology. What stays with people: the architecture, the decisions, and review of everything an agent produces.
Coming with an existing product? We start with a diagnostic, and the first answers arrive within a few weeks.
Development too.
We lead the strategy, research, product definition and design. When you want one team accountable all the way to a product in production, we build too.