All work
AdminArm enterprise data validation interface
AdminArmOriginal project work

Validation dashboard / processing state

Enterprise product / 01

AdminArm

Teams were manually comparing multiple Excel workbooks containing thousands of records, while revisions lived in a separate trail and long-running checks offered little shared visibility.

RoleLed product design - complete desktop experience
ScopeEnterprise data validation
CategoryEnterprise
Supporting project scale
  • 80+ production-ready screens
  • 50+ reusable components
  • 7 enterprise workflows
Read the story

01 / The problem

Teams were manually comparing multiple Excel workbooks containing thousands of records.

While revisions lived in a separate trail and long-running checks offered little shared visibility.

02 / Research & key insight

I found that comparison was only one stage in a larger lifecycle.

Mapping the work as a six-stage sequence revealed the need for continuity across processing, failure, revision, and recovery.

Project-supported findings & constraints

Supporting evidence: workflow mapping connected secure access, multi-workbook upload, automatic validation, cross-comparison, summarized results, and detailed discrepancy review.

  • 01Multiple Excel workbooks and thousands of records
  • 02Long-running validation and cross-comparison states
  • 03Revisions, incorrect records, export, and recovery paths

03 / Decisions & trade-offs

I structured the workflow as an explicit staged lifecycle.

Using prerequisites to unlock each action rather than exposing the whole workflow at once. Processing status, discrepancy review, revisions, and recovery remained visible as work moved forward.

My callClarity over immediate control density.

01

Progressive workflow

I revealed actions as prerequisites were completed, reducing decision fatigue and preventing accidental moves through the validation lifecycle.

02

Centralized validation status

I kept verification count, date, processing time, correct records, revisions, and incorrect entries visible without separate reports.

03

Tables built for scanning

I used color-coded states, fixed hierarchy, consistent spacing, and action-focused controls to support large-dataset review.

04 / Solution / workflow

The designed system

I turned that decision into Upload → Validate → Cross-check → Resolve. Centralized validation status, scan-friendly tables, revision history, discrepancy review, recovery paths, reporting, export, and cloud upload all follow the same staged logic.

  1. 01Upload
  2. 02Validate
  3. 03Cross-check
  4. 04Resolve
AdminArm workbook validation workflow screens
AdminArmOriginal project work

Staged workflow / upload to cross-check

05 / Outcome & evidence

24%reduction in validation time

50min → 38min. 

I turned manual workbook comparison into one guided lifecycle.

Keeping prerequisites, processing status, discrepancies, revisions, and recovery visible from upload through export. Documented project outcome: one centralized workflow for uploading, validating, comparing, reviewing, and exporting enterprise datasets.

06 / Reflection & learning

I learned that a high-volume validation result is not trustworthy.

Without its processing state and revision context. I now treat prerequisites and recovery paths as part of the main workflow, not secondary guidance.

Back to all work

Let’s make sense of it

A complex product.
A useful conversation.

I’m open to product design roles, consulting and selected collaborations.