Skip to content
All work

01 /2026

Real project

Product designer workspace

A tool where research, flows, screens and decisions form one chain — and every decision has a “why”. The case studies on this site are published from it.

Client
Own product
Role
Product, design & development
Year
2026
Status
In progress, 2026
Stack
Next.js · Supabase
Team
Solo, code with an AI assistant
Platforms
Desktop, tablet, phone
Project overview: nine stages with progress and a “What to do next” list
Project overviewWork · 01 / 06

There is no live site — browse the project pages below. Project pages ↓

How it works

A recording of the tool with the demo project, no sound.

A new project and its brief, a quote into a pattern, an insight with its sources, search to a decision with evidence. 42 seconds

Project pages

Click a page to view it full screen. Flip through with the arrow keys or a swipe.

On a phone

Page 1 of 17 · Project overview

01 /09

In short

Challenge

Design artifacts live in different places: interviews in documents, boards in Miro, mockups in Figma, decisions in chats. A month later nobody remembers why a screen looks the way it does, and the portfolio has to be pieced together from fragments.

Solution

One tool for the whole chain: brief → competitors → research → synthesis → insights → opportunities → flows → screens → decisions. Links between artifacts appear on their own when one grows out of another, and a finished project is published to this site as a case study.

Outcome

The tool is in use: I run my projects in it, and the case studies on this site are published from it. From any decision you can walk back to the quotes it stands on.

02 /09

How I worked

research screens: plan, guide, interviews, matrix
7research screens: plan, guide, interviews, matrix
levels: quote → pattern → insight
3levels: quote → pattern → insight
views: cards, matrix, UX review
3views: cards, matrix, UX review
impact and effort matrix
3×3impact and effort matrix
step types in the flow editor
8step types in the flow editor
required states: loading, empty, error
3required states: loading, empty, error
log format: why and what was rejected
ADRlog format: why and what was rejected

03 /09

Insights

INS-001

The “why” is lost first

The mockup outlives the project, the reasoning does not. When someone asks a month later why it is so, the answer is dug out of chats or made up again.

My own practice, project retrospectives
INS-002

Nobody maintains links by hand

Links between documents last until the first rename. Traceability works only if it appears by itself, as a by-product of the work.

Attempts to keep traceability in documents and spreadsheets
INS-003

The portfolio is pieced together from fragments

A case study is written after the project, from memory. If the process were already structured, the case would be a snapshot, not a separate job.

Preparing the case studies for this site
A quote is dragged into a column, the column becomes an insight: every card becomes its source

04 /09

Competitors

I compared not the products as a whole but how they cover the chain from research to decision. Each is strong in its own link, but the chain breaks at the hand-offs.

WorkspaceDocuments (Notion)Boards (Miro, FigJam)Research repository (Dovetail)
Quotes → insights with sourcesYesPartlyPartlyYes
Decision linked to evidenceYesPartlyNoPartly
Flow editorYesNoYesNo
Screen spec with statesYesPartlyPartlyNo
Decision log with rejected optionsYesPartlyNoNo
Project readiness by stageYesNoNoNo
Red — a chance to do better

05 /09

Opportunities

  1. OPP-01

    How might we keep the “why” together with the decision instead of in a chat?

  2. OPP-02

    How might we create links between artifacts without manual work?

  3. OPP-03

    How might we turn a project into a portfolio case study in one step?

06 /09

Flow

The main path: from an empty project to a case study on the site. Every step leaves an artifact the next one grows from.

  1. 01New project — one field
  2. 02Brief and competitors
  3. 03Interviews → quotes
  4. 04Synthesis board
  5. 05Insight → opportunity
  6. 06Flow and screens
  7. 07Decision with evidence
  8. 08Case study on the site

Edge cases

  • Empty workspace — an “Open demo project” button with ready-made data
  • Double click on the demo — one project, not two
  • Deleted artifact — links do not break, the panel shows what is gone
  • Forgotten password — recovery by email
  • Data loss — a JSON backup of the project
  • Phone — the flow canvas becomes a list of steps, tables scroll sideways with the first column pinned
A new project is one field; the brief saves itself and readiness grows

07 /09

Screens

Key screens of the tool. Each has the states that are easy to forget: empty, loading, error. Each also works on a phone: screenshots of both versions are in Project pages above.

  • Project overview on a phone: sections in a row at the top, 78% readiness and the stages

    Project overview

    Progress by stage and next steps — exactly what is unfinished

    • New project
    • In progress
    • All done
  • Synthesis board on a phone: columns scroll sideways, a card moves from its menu

    Synthesis board

    Quotes and observations are dragged into pattern columns; a column becomes an insight

    • Empty
    • No pattern
    • Ready for an insight
  • An insight's Links panel on a phone: sources, the chain and what it turned into

    Links panel

    Sources, the chain and what the artifact turned into

    • No links
    • Direct
    • Through the chain
  • A flow on a phone: a list of steps with branches and errors instead of the canvas

    Flow editor

    8 step types, branches and edge cases attached to steps

    • Empty canvas
    • Step selected
    • Not thought through
  • A decision's evidence on a phone: pain points, insights and user quotes

    Decision log

    Context, decision, why, rejected options and evidence

    • Draft
    • Accepted
    • Superseded

08 /09

Decisions and why

DEC-001Links are created automatically

Why: When an insight is written from a board column, the quotes in that column become its sources. Nobody keeps manual traceability (INS-002), so it has to be a by-product of the work.

Rejected options: Tags and manual links; A separate traceability diagram. Evidence: INS-002, OPP-02.

DEC-002A case study is a snapshot of the project, not a separate document

Why: Publishing freezes the project data at that moment. Later work does not reach the site until the case is published again, and the portfolio does not have to be written from scratch.

Rejected options: A separate CMS for the site; A live view of the project without a snapshot. Evidence: INS-003, OPP-03.

DEC-003A demo project instead of an empty screen

Why: An empty tool explains nothing. The demo shows the whole chain on an example, and it is safe to break.

Rejected options: An onboarding tour with tooltips; An empty screen with instructions. Evidence: First sessions in the tool.

DEC-004Access by invitation

Why: This is a working tool with client data. Sign-up is open only to addresses on a list, and every row in the database is protected by access rules.

Rejected options: Open sign-up. Evidence: Security audit before launch.

DEC-005The phone is a full version, not a viewer

Why: A project has to be opened and fixed where it is needed: in a meeting, on the go. So the phone gets the same tool: sections in a row at the top, the flow canvas as a list of steps, wide tables scrolling sideways, and search that reaches any artifact. Every release checks the pages on a phone screen automatically: nothing may run off the edge.

Rejected options: A read-only version; A separate mobile app. Evidence: Automated tests on a phone screen, Phone screenshots above.

Search leads to a decision, and the decision to the pain point, insight and quotes it stands on

09 /09

Results

The tool is launched and I run my own projects in it. This site is its continuation: case studies are published from the tool, and the pages of the site and the tool are checked automatically for accessibility and phone layout.

stages in the project chain
9stages in the project chain
tables with row-level access rules
46tables with row-level access rules
automated database checks
276automated database checks
site languages: Ukrainian and English
2site languages: Ukrainian and English

Next case · 02 / 06

Neural WebCam

Open case

Tell me briefly about the product and the timeline — I'll reply within a day.