First get a clear picture of material, context, roles and limitations.
Process / method
From demand to live system
Six concrete steps connect the business question to data, system logic, interface and human control. AI accelerates analysis and construction, but the product choices, evidence limits and final control remain mine.
Context first
Choices explicit
Output presentable
Relevance clear
Determine which form is most usable and professional.
Not only deliver, but also show why it works.
Six-step route
From context to live
Each step has its own artifact, an explicit decision point and a check before work continues.
Demand and environment
User, goal, risk, precondition and desired outcome become clear.
Sources and boundaries
I determine what is usable, traceable, permitted and still uncertain.
Route and exceptions
Rules, scenarios, states and escalations form the system contract.
Action and understanding
The user sees what is needed to choose, correct or continue.
Proof and agreement
Source, uncertainty, status and human decision-making power remain visible.
Test and delivery
The proof becomes operable, responsive and honest about what has and has not been proven.
Working principle
First the gist
As a result, the outcome does not remain superficial. First it must be clear which system, which story or which user problem is actually on the table. Only then does the form follow: interface, workflow, visuals, text or presentation.
Artifacts and decision moments
What is actually created per step
Not an abstract creative process, but material with which a team can test, decide and build further.
Context & data
- problem definition and stakeholder map
- source and data contract
- risks, assumptions and acceptance criteria
Decision moment: is the demand sharp and is the input responsible enough to build?
Logica & interface
- states, rules and exception routes
- flow, wireframe and working interaction model
- example scenarios with expected outcome
Decision moment: can a user understand what the system does and why?
Controle & live
- approval gates and audit trail
- test scenarios, responsive QA and error states
- status, limitations and next proof step
Decision moment: is the proof safe, fair and strong enough to be tested by others?
AI as a construction partner
Fast and controlled
I use AI for research, variants, code, analysis and test support. I determine the product direction, select sources, test output, monitor human control and visibly mention when a proof uses test data, simulation or as yet unproven assumptions.
Operational map
Targeted adjustments
This process is not linear on paper, but functions as a control loop: reading signals, tightening course, building, looking back and adjusting again.
The first step is almost always to reduce chaos to a bounded problem, a clear question or an actionable ambition.
- Gathering context
- Naming roles
- Making limitations visible
Not every problem requires the same output. Sometimes that is a system, sometimes an interface, sometimes a brand layer.
- Determine structure
- Choose priority
- Design reading route
The implementation remains iterative: testing, tightening, rearranging and testing again until the outcome is both smart and useful.
- Make it work
- Make it presentable
- Validate and adjust
Method
From source to result
This is the solid backbone behind the projects. Not every project uses exactly the same resources, but the logic remains the same.
Context
Make the project name, domain, role, reason and limitations short and sharp so that the starting point becomes clear.
Structure
Decide which form fits best: system, flow, interface, brand layer or combination thereof.
Execution
Build, write, design, test and rearrange until it not only sounds smart, but also works.
Relevance
Translate into what an employer, client or user actually benefits from it.
Process in work
Visible in cases
Not as an abstract method map, but as concrete traces in different types of projects.
Analysis to system
HoiJob
From vacancies and rich profile material to an explainable AI system with traceable output.
Direction to surgery
Phoenix Garden
From a separate restaurant problem to a coherent concept with brand, routing and team logic.
Experiment to pipeline
AI Animation & Audio-to-Face
From creative test to a pipeline in which voice, image, timing and validation come together.
Interface as system
Star Citizen Universe
Show how interface, content and community patterns come together in a recognizable digital world.
Collaboration effect
Practical result
The method is especially useful when many disciplines move at the same time and a sharp outcome is still required.
Less searching time
Because the process forces core choices and clear structures at an early stage.
Fewer misunderstandings
Design, technology and content remain better aligned when choices are made visible.
Better presentable work
Important for portfolios, decision-making, teams and employers who want to see how someone really works.
Next step
View the cases
The work page contains the entire case index, so that the working method does not remain abstract but is linked to real material, real choices and real output.