Process

Process

A Way to Turn a Complex Product Task into a Clear Interface System

From context and scenarios to UX architecture, components, visual layer, validation, and handoff.

A systematic end-to-end method for interface design: from decomposing product logic to runtime validation and specified design delivery.

From context and scenarios to UX architecture, components, visual layer, validation, and handoff.

Key principles

01.

Context Before Interface

Before the visual layer, it is necessary to understand the domain: who the users are, what roles they have, which tasks they solve, which entities participate in the product, and where the real constraints are.

02.

Routes and IA

The interface is designed as a system of transitions, not a set of separate screens. It is important to see entry points, user flows, screen families, return paths, empty states, and context recovery scenarios in advance.

03.

Component Decomposition

Product logic should be decomposed into clear levels: foundations, Product UI Kit, Product Components, Component Blocks, Page Layouts, and runtime adapters.

04.

Visual Layer as Structural Support

Typography, grid, color, spacing, states, and responsive density should help the user understand the system, not mask unresolved logic.

05.

Validation and Handoff

Solution readiness is defined not only by how it looks. States, responsiveness, data, component boundaries, documentation, and how the solution will be handed over to development need to be checked.

How the Process Works

The work does not begin with final screens, but with how the product is structured: roles, entities, data, states, constraints, and decision points. After that, product logic is translated into user flows, screen families, a component model, visual language, and materials for implementation.

This approach is especially useful in B2B, B2C, and enterprise products where the interface must withstand growth: new roles, scenarios, data types, states, implementation constraints, and public / private boundaries of materials.

It is not a linear scheme of “complete a stage and forget it.” In real work, some decisions are refined iteratively. But the overall order helps avoid starting with the visual layer when the product logic has not yet been analyzed.

Phase 1. Context and Task Frame

The first stage is to understand the product and find the real complexity.

At this level, it is important to answer several questions:

  • who the users are and what roles they have;
  • which tasks they solve;
  • what counts as a successful result for them;
  • which scenarios are primary and which are supporting;
  • which entities and data define the interface;
  • where the user makes decisions;
  • which constraints exist from the business, team, implementation, or data.

The result of this phase is a product frame: roles, scenarios, entities, key constraints, and an understanding of where the main interface complexity lies.

Phase 2. Routes and Information Architecture

After the context, the product is decomposed into routes.

Here it is important to see not only separate pages, but also how the user enters the product, moves between scenarios, where context can be lost, and how they return to work.

The work includes:

  • entry points;
  • user flows;
  • navigation model;
  • information architecture;
  • screen families;
  • search / catalog behavior;
  • detail page structure;
  • transactional flow;
  • account-side continuation;
  • empty, error, and recovery states.

For example, a marketplace product may start with discovery, move into search/catalog, then to a detail page, offer selection, booking flow, confirmation, and account-side scenarios. An enterprise product may start with a workspace, move into a repository, detail screen, evaluation flow, dashboard, or organization layer.

At this stage, the logic of movement through the product appears, not just a list of pages.

Phase 3. UX Architecture

UX architecture translates the product model into an interface structure.

At this level, decisions are made about:

  • which entities become pages;
  • which remain filters;
  • which are the offer / inventory layer;
  • which elements should be badges;
  • which blocks belong on a detail page;
  • where the transactional entry point appears;
  • where an account gate is needed;
  • which states are a normal part of the product, not edge cases.

This is especially important in complex systems. If a beach, commercial venue, offer, filter, and badge look similar but have different roles inside the product, the interface should preserve that difference. If a requirement in an enterprise system is connected to a project, vendor, evaluation, and discussion, it cannot be designed as a simple row in a table.

UX architecture is needed so the user sees a simple surface while the system underneath remains logically clean.

Phase 4. Component System

When scenarios and product structure are clear, the interface is decomposed into components and levels of responsibility.

Several layers usually appear in the work:

Foundations

Tokens, color, typography, spacing, radius, surfaces, shadows, and basic visual rules.

Product UI Kit

Base interface elements: buttons, fields, badges, chips, controls, overlays, menus, navigation primitives, feedback states.

Product Components

More meaningful elements: cards, offer rows, reservation rows, notification rows, payment method rows, filter units, account tiles.

Product Component Blocks

Large reusable sections: search toolbar, filter panel, hero media header, detail support blocks, BookingWidget, account sections, recommendation rails.

Product Page Layouts

Page assembly rules: section order, grid, responsive behavior, sticky zones, mobile stacking, sidebar behavior, and conditions for showing blocks.

Runtime adapters / View-model layer

The layer that prepares data and states for the UI: routes, filters, selected offer, map/list state, booking draft, source-backed labels, callbacks.

This separation helps avoid mixing visual anatomy, business logic, runtime state, and page-local decisions.

Phase 5. Visual Layer

The visual language appears after the product structure is understood.

This does not mean visual design is secondary. On the contrary, a good visual layer helps the user understand the system faster: where the main scenario is, where a secondary action is, where a status is, where a filter is, where an error is, where a restricted state is, and where a transactional step is.

At this stage, the work includes:

  • typography;
  • spacing rhythm;
  • grids;
  • color semantics;
  • surface system;
  • card hierarchy;
  • badge / chip system;
  • focus, hover, active, disabled states;
  • empty and fallback states;
  • responsive density;
  • dark / light or brand-specific logic, if needed.

Phase 6. Runtime Validation and Reference Surfaces

Some interface decisions are better validated not only in a static mockup, but also in a working environment.

This can be a Live Demo, showcase app, interactive prototype, component reference surface, or local React / Vite / Next.js assembly.

Runtime validation helps reveal:

  • how layout reacts to different data;
  • how cards work with long names or without photos;
  • how map/list sync is connected to card states;
  • how overlay, sheet, and sticky zones behave on mobile;
  • how selected offer gets into the BookingWidget;
  • how empty / no-photo / no-results states look without manual adjustment;
  • where a separate component is needed and where adapter logic is enough.

Reference surfaces are needed to validate components outside the user journey. They make it possible to view states, variants, cards, filters, booking units, account components, responsive frames, and fallback scenarios separately from the final page.

Phase 7. Documentation and AI-assisted Workflow

Documentation in this process is not an archive after the project, but a working layer.

It helps:

  • capture the product model;
  • separate source of truth from temporary notes;
  • explain component boundaries;
  • preserve decisions about screen families;
  • guide AI agents and developers;
  • choose the right checks;
  • prepare handoff.

AI can help with analysis, inventory, option preparation, public-safe copy, finding contradictions, documentation updates, and QA notes. But it should work within rules: first the task layer, then source docs, then a small work slice, validation, and human review.

This way, AI speeds up the process but does not replace product responsibility.

Phase 8. Validation and Handoff

The final stage is to verify that the solution is truly ready to be handed over.

Different task types require different checks:

  • a docs-only change is checked for coherence and source-of-truth;
  • component change — for states, variants, API / props, and reference surface;
  • runtime behavior — for build, typecheck, browser QA, responsive behavior, and edge states;
  • visual change — through human visual QA;
  • architecture change — for impact boundaries and adjacent product layers.

Handoff should explain not only “how it looks,” but also “how it works”:

  • what changed;
  • why it was done;
  • which layer was affected;
  • which components or documents are connected;
  • which checks were completed;
  • what was not checked;
  • which limitations remain;
  • where human review is needed.

This allows the solution to continue evolving without guesswork or rebuilding context from scratch.

What This Process Provides

This process helps design complex interfaces as resilient product systems.

It reduces the risk that:

  • identical patterns will be solved differently;
  • Figma and code will diverge;
  • a temporary solution will become permanent;
  • a component will lose its boundaries;
  • visual polish will break layout;
  • the product model will mix with UI labels;
  • handoff will require constant clarification.

The main value of the process is manageability. It helps move from complex product logic to an interface that can be explained, validated, assembled, handed over to development, and evolved further.