Find the Needle,
Skip the Haystack
When Every Detail Matters: Designing a patented UX for high‑stakes legal review
Company
Everlaw (Legal Tech)
Role
Senior Product Designer
BACKGROUND
One difference changes everything
To prepare cases, legal teams review tons of documents. They often face clusters of documents that look nearly identical. However recognizing one small difference could be the key to winning a case.
The work is high-stakes and deadline-driven, but the workflow is traditionally manual: compare versions, scan line-by-line, and hope you don’t miss the one line difference that changes everything.
PROBLEM
Drowning in near duplicates
Manually scanning similar-looking documents is a tedious, expensive, and error-prone process that only gets more complicated as the pile of documents needing review grows
Traditional side-by-side comparison cannot scale
Example of traditional two-document comparison
In our analysis of document groups, over 28% of near-duplicate documents have 10 or more documents.
So, the more similar documents there are to review, the higher the cost and higher the risk of missing key information.
When everything looks the same, it’s easy to miss what matters
A single word can change the outcome of a case—but in document review, finding that change often means re-reading the same near‑duplicate text across dozens of documents.
And even when a reviewer finally spots the difference, the work isn’t over: they still have to apply consistent coding to the right subset of documents, quickly and without errors.
Spotting the difference between near-duplicates
MY APPROACH
Designing a clear path forward
This was a large, ambiguous problem without an existing pattern we could enhance or replicate. The existing pattern of two-document comparison fundamentally failed for large near‑duplicate document sets.
Defining the real problem
Setting scope & success criteria
Iterating effectively
I started by grounding the work in the legal reviewer’s reality: they’re under time pressure to find winning evidence.
Every extra minute spent re-reading duplicates is time taken from higher‑value legal work.
From that lens, I defined the core jobs-to-be-done and used them to align our team on the goals:
surface meaningful differences without re-reading
stay oriented across versions
apply decisions immediately to the right documents
I facilitated working sessions with PM and engineering to define a clear scope and direction.
We explored multiple models, pressure-tested feasibility early, and iterated quickly—treating engineering as a thought partner from ideation, not a downstream handoff.
This process produced our guiding question:
How do we help legal reviewers quickly identify and act on meaningful differences across near‑duplicate documents?
SOLUTION
Difference Viewer:
Find the Change That Matters—Fast
Helping legal teams increase efficiency and move cases forward with confidence
A difference viewer like no other
Faster review. More consistent decisions. Stronger case outcomes.
Difference Viewer allows users to quickly see the differences of many similar documents at once:
The differences are shown in boxes
When a difference is selected, all variants are shown
This view allows users to read the similarities once and compare all the differences in one interface.
EFFICIENT
Review the whole set of docs—all at once
Difference Viewer brings an entire near‑duplicate group into one view—so reviewers stop bouncing between pairs and start understanding the set: what’s shared, what diverges, and where to focus.
Spot the change without the hunt
Repeated text fades into the background while differences are clearly highlighted.
A variants panel summarizes each change and lets reviewers jump straight to it—turning “searching” into “seeing.”
Take action in seconds
Once a critical change is identified, reviewers can code directly in the viewer.
Apply a decision across the full group when it’s the same, or target only the documents with a specific variant—so teams move faster and reduce mistakes.
ACCURATE
IMPACTFUL
Outcomes
Giving Everlaw’s users and Everlaw as a company, an edge with a patented UX.
I was named an inventor on US 12,265,787 B2 (“Document Difference Viewing and Navigation”, Everlaw, Inc.) covering the system and interaction model for multi-document difference viewing, navigation, and variant-based actions.
“Variant-based review helped the team stay consistent across duplicates.”
Reduced time-to-decision on near-duplicate sets by up to 87%
“We can review a whole cluster of near-duplicates in one pass instead of re-reading the same document over and over.”
“Differences are obvious immediately—no more side-by-side hunting for what changed.”