An AI-native design and product partner

Come to us before the problem is fully defined.
Business and product design, UI/UX, AI experience design, engineering and validation.

Talk to us about your projectTalk to us about your projectSee how we think
Connecting craft with evidence, design with engineering, people with AI — we design the way your organisation solves problems.
01 — Issue

AI has made design, writing and code far faster to produce.But speed delivers nothing while what you should be building stays vague.

  1. 01

    The real problem
    has never been framed

    There are requests, but the problem
    underneath them is never put into words

  2. 02

    Business, design and engineering
    decide in isolation

    Every team starts from different assumptions,
    so each decision reopens the same debate

  3. 03

    AI adoption stops
    at buying the tools

    The tools are in place, but they never reach
    the workflow or the product itself

  4. 04

    The UI improves
    but the business doesn't

    The screens look better, yet nobody can say
    whether a metric actually moved

  5. 05

    Nothing is left behind
    once the project ends

    When the outside partner leaves,
    improvement stops with them

  6. 06

    Output went up,
    quality and coherence went down

    AI produces more of everything, while
    the brand loses its consistency

02 — Approach

We are not a studio that simply builds what it is handed.We work across business, users and technology to find the problem actually worth solving.

Research, strategy, UX, UI, prototyping, engineering and measurement are designed as a single flow — so that with every project, your organisation decides and executes faster.

03 — Blueprint

How customer input, AI and codified know-how (Skills) connect into a single lap.This is our business and learning loop, drawn exactly as it runs.

HAVU / SYSTEM BLUEPRINT 01

The design studio for the AI era — business & learning loop

Customer input becomes organisational knowledge through Claude and Skills, then returns to the next design.

01 — ACTORS

  • DESIGN STUDIO

    Who builds

    UI / UX · Communication · Product · Graphic

  • CLIENT

    Who decides

    Business problems · Brand · Decisions

  • CUSTOMER

    Who uses

    Experiences it, and answers back

02 — OPERATING & LEARNING LOOP

CLAUDE

AI PARTNER

Analysis, generation, improvement

SKILLS.MD

Codified know-how

Standards, steps, history

  1. VOICE

    Customer input

  2. KNOWLEDGE

    Becomes know-how

  3. DESIGN

    Feeds the next design

  4. DELIVERY

    Built and shipped

  5. EXPERIENCE

    Becomes experience

03 — ACCUMULATION

It compounds every lap

The more we run it, the sharper Skills

and design quality become.

HAVU's business and learning loop: customer input is turned into organisational knowledge through AI and Skills, then fed back into the next design.

Customer input becomes organisational knowledge through AI and codified design know-how (Skills), then returns to the next design. The longer it runs, the better the quality gets.
04 — VALUE

These are the five things we deliver.

01

Find the right problem

We bring together user research, business requirements, data and frontline input, and structure the problem sitting behind the stated request.

02

Make hypotheses tangible, fast

We use AI across research, idea generation, prototyping and engineering to reach something testable quickly.

03

Move people

Beauty, surprise, warmth and a distinct brand voice don't come out of logic alone. We shape them with human judgement and a designer's eye.

04

Carry it through to build and run

We don't stop at handing over a design. We design the engineering, the design system, the AI workflow and how the team runs it.

05

Turn results into learning

We combine quantitative data with qualitative research to confirm what worked and what didn't, and feed it into continuous improvement.

Stance

How we work with AI

We use AI to widen exploration and increase speed — not to replace human judgement. Final calls on quality, context, ethics and brand stay with people, who remain accountable for them.

05 — Services

You don't have to hand us the whole thing.We combine only the parts you need, around your goals and your existing team.

01Product & Business Design

Turn a business concept into a product you can actually ship

We convert business concepts and half-formed ideas into products with real user value and a realistic path to delivery. Rather than building the finished thing up front, we identify the uncertainties that matter commercially and test them in small steps.

  • New business and new product design
  • Problem definition and value proposition design
  • User research
  • Customer journeys
  • Service blueprints
  • Requirements and roadmap design
  • KPIs and validation planning
02UI/UX Design

Design experiences that move the business, not just usability scores

We start by agreeing which user behaviour and which business metric should change, then work backwards into information architecture and UI. The reasoning behind each design decision is recorded and shared — including the parts decided by instinct.

  • UX audits and problem analysis
  • Information architecture
  • Wireframes
  • UI design
  • Prototyping and user testing
  • Accessibility improvements
03AI Experience Design

Design people and AI working together — not an AI feature bolted on

We define whose problem is being solved, where a human makes the call, and how it fits into the workflow or the product. We assume it will sometimes be wrong, and design the evidence, the checks and the handover to a person accordingly.

  • UX design for AI products and AI features
  • Conversational and generative UI design
  • Workflow design around AI adoption
  • Human-in-the-loop design
  • Trust, explanation and error design for AI output
  • AI workflow and Skills design
04Design Engineering

Close the gap between design and code so the product actually moves

We align component structure, naming, design tokens, states and responsive rules to give designers and engineers a shared language. We stay involved through implementation review and design QA.

  • Frontend engineering
  • Prototype development
  • Design system build-out
  • Component design
  • Design-to-code integration
  • Quality and performance improvement
05Design Organization & AI Transformation

Build the machinery for consistently good decisions

We leave behind decision criteria and process, not a one-off deliverable. The goal is something the whole team can reuse — not knowledge locked in one person's head.

  • Design process improvement
  • AI policy and guidelines
  • Standardising research and validation
  • DesignOps
  • In-house workshops
  • Knowledge, templates and decision criteria
06 — Process

Every project runs through these five steps.

  1. Step 01

    Understand

    We gather input from leadership, the business, the frontline and users, and agree on where you are and where you're going.

  2. Step 02

    Frame

    We organise the problem, hypotheses, constraints and success measures, and define the question this project has to answer.

  3. Step 03

    Explore

    We make several directions visible quickly and turn them into prototypes concrete enough to argue about.

  4. Step 04

    Build

    We build the chosen direction into a UI, a product and a workflow people can actually use.

  5. Step 05

    Learn

    We review usage, KPIs and user response, and keep it as knowledge that drives the next improvement.

07 — Difference

The goal is change, not a deliverable

How our work differs from conventional production support.

Conventional production supportHow we work
  • Builds what was requestedStarts from what should be built
  • Hands over a UI or a websiteDesigns a system that drives business results
  • Supplies designer hoursSupplies judgement, exploration and speed
  • Separates design from engineeringConnects concept, design and engineering
  • Builds it once and stopsMeasures, learns and keeps improving
  • Leaves the know-how outside your companyLeaves decision criteria and systems with the client
  • Uses AI to produce work fasterBuilds AI into how the organisation solves problems
08 — Legacy

Improvement shouldn't stop the moment an outside partner walks away.

The decision criteria, components, research findings, AI workflows and validation methods created during the project stay with your organisation in a form it can reuse.

A record of problems, hypotheses and decisions, Design principles, Design system, UI components, Research and validation templates, How KPIs map to UX, AI Skills, prompts and workflows, Operating guidelines, A roadmap for what to improve next
A roadmap for what to improve next, Operating guidelines, AI Skills, prompts and workflows, How KPIs map to UX, Research and validation templates, UI components, Design system, Design principles, A record of problems, hypotheses and decisions
09 — PHILOSOPHY

AI can generate answers by the thousand.
What it cannot decide is which question to ask.

We find the problem nobody has put into words yet,
and connect business, users and technology into one experience.

We don't retreat into data and rationale alone,
and we never let go of the craft that moves people.

We don't stop at the idea —
we stay until it ships, gets used, and produces a result.

And when the project ends,
we leave behind a way for you to keep thinking, building and learning.

That is what design in the AI era means to us.

Insight
Find the problem worth solving, hidden behind the stated request
Logic
Build the evidence for a decision from research, data and business needs
Sensibility
Refine the beauty, the friction and the brand voice that move people
Making
Give it real form as a UI, a prototype and a product
Learning
Use measurement and validation to make improvement continuous
10 — TEAM

We don't stop at strategy alone, design alone or engineering alone; we bring each discipline across the whole project. And rather than treating the client as the party placing an order, we work as one team around a shared question.

  • Kei Kawashima

    Kei Kawashima

    CEO, PM, Director / UI/UX designer, frontend

    Works across public- and private-sector business systems, education, e-commerce, apparel, corporate and entertainment web and app projects. Specialises in UI/UX design, frontend and coding.

  • Rion

    Rion

    Assistant, apprentice UI/UX designer / Frontend coder

    A 16-year-old still at public high school, working part-time as a frontend coder and currently deep in learning UI/UX design.

12 — FAQ

Get in touch even if the problem and requirements aren't settled yet.

Contact us

HAVU｜Design and product partner for the AI era