Discovery
We read your evals, model cards and support tickets, because the failure profile is the brief for most of the interface.
AI
Most AI products that fail, fail on trust
We design the interface that shows what the model did, so people can check it.
What we do
AI people can check, steer and trust
Multi-step behaviour, task handoffs, interruption and resumption, and the boundary between acting alone and asking first every time.
See FreelogoAI products we have already shipped
Shipped
Freelogo
An AI logo platform for HostPapa, with plain-language input, speed with control and one set of rules across thousands of logos. Concept to MVP in 3 months.
TSRGD
Search across 600 pages of law, with the regulation, the diagram and the project on one screen so the person using it can defend what they did.
Wise
Compliance checks surfaced where the decision is made: Right to Work, DVLA and CIS, document expiry and the pay run, each one checkable.
AI in a regulated setting
The same pattern every time: show what the system did, how sure it is, and how a person checks or overrides it. Reliability sets what it may do alone.
“The work speaks for itself.”
The process
Six stages over 12 weeks, to a date we hold
Decision boundaries
We draw the line between what the system does alone, what it proposes for approval, and what it refuses to touch.
Prototypes
Clickable, in front of users, running against real output rather than a screenshot of a good day.
The trust layer
Sources, confidence, what changed, what can be undone and what was logged, readable in seconds by someone busy.
Interface design
Production UI with every state designed. Streaming, partial, empty, stale, wrong, rate-limited and unavailable.
System and handover
A component library your engineers build from rather than around, documented so it holds when the model changes underneath.
Related services
FAQs
Questions we get
Designing the interface a person uses to work with a model: what the system does automatically, what it asks about, how the output is presented, how it is checked, and what happens when it is wrong. The model is one component. AI UX is everything around it.
Show the working and let people correct it. Trust in an AI product comes from the interface making it obvious when to rely on the output and when to look twice. We design the human-in-the-loop steps first, then the happy path.
Then we are the wrong studio. A feature built for show gets shipped, ignored and quietly removed a year later, after eating a quarter of the roadmap. That is a much cheaper conversation to have before the build than after it.
That is a design brief, not a blocker. An interface that communicates uncertainty honestly outperforms one that performs certainty, especially with technical buyers who can smell the difference. Reliability sets what the system may do alone. It does not decide whether you can ship.
No. We design the product around it, working alongside your ML or engineering team. We will read your evals and your docs rather than ask for a summary, and we will not pretend to opinions about your architecture that we have not earned.
Good. In a regulated product, transparency is what the buyer is paying for. Audit trails, reversibility, human sign-off and refusal states are all design work, and they are the parts of this we most like doing.
Twelve weeks for a full engagement, discovery through handover. A trust layer on an existing product is shorter. You get the date at proposal stage, before you commit to anything, and we scope to the actual gap rather than sell you everything.
Pick the number before we start. Usually feature adoption, completion rate, override rate or how often a user accepts the first output. We agree it in discovery and report against it after launch.
Related reading
Worth reading before you call
Nobody uses AI they cannot check
Almost right is the expensive kind of wrong. It moves the work from doing the task to checking whether the machine did it, and most products have not designed that part at all.
AI disclosure is a design problem
Since 2 August 2026 the EU AI Act requires you to tell people when they are dealing with a machine. Most teams will answer that with a footer, which is the one place nobody looks.
AI sends fewer, better visitors
AI assistants send about one visit in seven hundred, and those visits convert better than anything else you have. That changes what a B2B website is for.