Project Summary
Project: Building an Evidence Mapping Workflow for Legal Teams
Role: Senior Product Designer
Timeline: 6 weeks
Team: Product Manager, Applied Scientists, Engineers
My Contribution
Took ownership of an existing initiative after inheriting the project
Synthesized prior research and customer feedback
Led end-to-end UX and interaction design
Built production-ready prototypes
Accelerated implementation using Cursor, Claude, and GitHub
Designed and developed MVP within six weeks
Overview
Evidence Mapping helps litigation teams connect legal claims with the evidence needed to prove or refute them. Although this workflow is one of the most valuable parts of preparing a case, attorneys typically manage it manually using Word documents and spreadsheets.
The goal of this project was to bring this workflow into our platform by combining AI-assisted analysis with human review, creating a living artifact that evolves throughout the lifecycle of a case.
Rather than replacing attorney judgment, the system was designed to accelerate the tedious work of organizing evidence while keeping attorneys in complete control of every decision.
The Challenge
I inherited this project after initial discovery had already begun.
Before designing solutions, I needed to quickly understand:
- Existing research
- Legal workflows
- Previous product decisions
- Technical constraints
- Emerging AI capabilities
I reviewed customer interviews, attended user group meetings, partnered closely with Applied Scientists, and immersed myself in how litigation teams actually build evidence maps.
At the same time, the project carried an aggressive deadline—our team had roughly six weeks to design and build the first version.
Understanding the Problem
Research consistently showed that evidence maps are among the most valuable strategic artifacts attorneys create, yet the process remains highly manual.
Teams were:
- Building evidence maps in Word or Excel
- Copying evidence between disconnected systems
- Manually searching for supporting documents
- Struggling to keep charts updated as cases evolved
The result was a workflow that was time-consuming, difficult to maintain, and disconnected from the rest of the case lifecycle.
What I Learned
Through user interviews, customer advisory sessions, and collaboration with Applied Scientists, several themes consistently emerged.
Evidence maps are not one-size-fits-all
Attorneys rarely build an evidence map from the beginning of every case.
Instead, they create focused evidence maps for:
- Summary Judgment
- Trial preparation
- Mediation
- Complex investigations
- Early Case Assessment
This shifted our thinking away from designing a rigid workflow toward a flexible workspace.
Research must be controllable
Attorneys wanted AI to assist—not overwhelm.
They needed to:
- Choose which documents to analyze
- Control search scope
- Prioritize research
- Understand the cost of broader searches
Trust increased when users remained in control.
AI-generated facts require verification
Users viewed generated facts as starting points rather than final answers.
They wanted to:
- Review every fact
- Edit generated content
- Approve evidence
- Track changes
- Understand why AI made recommendations
This reinforced that human validation needed to be central to the experience.
Design Principles
From the research, I developed several principles that guided every design decision.
AI should accelerate—not automate—legal reasoning.
Users should stay in control of every important decision.
The workflow should remain flexible enough to support different litigation strategies.
Every AI-generated output should be transparent, editable, and reviewable
The Solution
The experience begins by allowing attorneys to upload one or more seed documents, such as a complaint.
AI analyzes the documents and generates:
- Claims
- Legal elements
- An initial research plan
Rather than immediately searching every document in the workspace, users first review and refine the generated outline.
Once approved, the agent analyzes the selected document set and constructs the initial evidence map.
Each legal element adds a layer where attorneys can review evidence, validate findings, and strengthen their legal arguments.
Evidence Mapping
For every legal element, users can review:
- Supporting facts
- Conflicting evidence
- Confidence scores
- Citations
- Document references
- Dates
- Source information
Each fact could be:
- Approved
- Edited
- Pinned
- Commented on
- Deleted
This created a collaborative workflow where AI handled extraction while attorneys remained responsible for legal judgment.
Flexible Research
One of the largest design improvements was making research iterative.
Users could:
- Edit the research plan
- Add or remove legal elements
- Analyze saved searches
- Rerun AI after revisions
- Customize map columns
- Sort and filter findings
- Instead of forcing users into a rigid workflow, the system adapted to different litigation strategies.
Designing for Trust
Because this feature relied heavily on AI, trust became one of the primary UX challenges.
To improve confidence, I emphasized:
- Editable AI output
- Visible confidence levels
- Citations for every generated fact
- Document traceability
- Manual approval workflows
- Transparent research plans
Rather than presenting AI as an authority, the interface treated it as an intelligent assistant.
Building with AI
This project also transformed how I worked as a designer.
Using Cursor and Claude alongside our GitHub repository, I was able to move beyond static mockups.
I built production-ready prototypes, explored interaction patterns directly in code, and collaborated more closely with engineering throughout implementation.
Working this way dramatically shortened the feedback loop between design and development and allowed the team to ship within six weeks.
Outcome
The MVP introduced a fully integrated evidence mapping workflow inside our platform, replacing fragmented spreadsheet-based processes with an AI-assisted experience.
The feature enabled attorneys to:
- Map claims to evidence
- Identify gaps earlier
- Organize legal strategy collaboratively
- Keep evidence connected throughout the case lifecycle
The project also established a foundation for future capabilities, including reusable evidence maps, broader investigation workflows, and AI-assisted legal research.
Final Thoughts
This project reinforced that successful AI experiences are rarely about automation alone.
The biggest design challenge wasn't generating information—it was designing the right balance between AI assistance and human judgment.
It also changed how I approach product design. By combining traditional UX methods with AI development tools like Cursor and Claude, I was able to contribute not only as a designer but also as a rapid prototyper working directly alongside engineering.
The result was a feature that shipped quickly, solved a meaningful customer problem, and laid the groundwork for a broader AI-powered case strategy platform.