CASE STUDY

Checklist AI / Verification Assistant

An AI-assisted verification workflow that turns complex document review into structured, traceable findings — while keeping evidence and reviewer judgment close to every decision.

Role Product Designer
Focus Verification UX
Tools Figma
Checklist AI verification interface showing a completed document review and an actionable finding

Project Context

Document review needed one connected flow

Document-heavy review gets slow when requirements, sources, findings, and decisions live apart. This feature brings them into one path from checklist and document to findings users can inspect and act on.

The UX Challenge

Make verification clear enough to trust

  • Keep long checklists and documents understandable as a review flow.
  • Show progress without exposing internal AI reasoning.
  • Make every finding traceable to requirement, source, and evidence.
  • Let reviewers challenge results without losing context.

My Role

Designing the verification journey

I designed the end-to-end verification experience: hierarchy, traceability, review states, and the relationship between AI assistance and human judgment.

Interaction Model / Core Flow

A structured path from document to decision

The workflow moves from review setup through verification to a human decision—keeping requirements, findings, and evidence connected.

Verification workflow showing Set Up the Review, Verify, and Decide

Along the way, reviewers can inspect evidence, challenge a finding, ask follow-ups, and prepare report content before confirming an outcome.

Key UX Decisions

Designing for trust and review confidence

Structured findings

Present verification results as inspectable findings rather than long AI-generated answers.

Evidence stays connected

Keep the requirement, source location, evidence, and finding connected so users can verify results quickly.

Visible review progress

Show meaningful review stages without exposing unnecessary internal AI reasoning.

Context follows the user

Actions and follow-up questions retain the active finding, requirement, source, and evidence.

Human judgment stays explicit

AI can explain, compare, suggest, and prepare content, but the reviewer remains responsible for important conclusions.

Design System / Interface Logic

Consistent patterns for compliance-critical tasks

Findings, requirements, source evidence, and reviewer actions share one interaction structure so each issue remains inspectable and challengeable.

Verification workspace connecting review findings, finding details, and source evidence

The interface separates what the system found from the evidence behind it and the decision the reviewer still needs to make.

Final Design

A verification workspace built around review

Review setup, findings, evidence, AI assistance, and reporting sit in one verification path: finding → requirement → evidence → reviewer decision.

The interface does not ask people to trust a generated conclusion on its own. Each finding links back to its requirement and source evidence, then forward to the next action.

AI helps organize and investigate the evidence. The reviewer remains responsible for the decision.

Design Outcome

Finding, evidence, and judgment stay connected

The review pattern can be reused across different document types because the finding, supporting evidence, and reviewer decision always stay connected.

See something you’d challenge?

I like talking through design decisions. If you would have approached something differently, I’d like to hear it.

Next project

Explore another project

See how I approached a different product problem.