CLIENT
Product
IMPACT METRICS
2+
refactored products
1k+
components created
19%
increase in engagement
1
design system
30+
fragmented patterns adjusted
90-95%
task completion
MY ROLE
Senior Product Designer
TEAM
Senior Product Designer
Head of Design
5+ Developers
1 Project Manager
TOOLS
Figma, FigJam, Miro
LAUNCHED IN
Design System /
1 month
ROLE
Senior Product Designer
When I joined ContractPodAI, I was working on Leah, an AI-driven contract lifecycle management platform used across multiple products in the same ecosystem.
The challenge wasn’t just UI inconsistency, it was that the product had grown quickly without a clear structure.
Workflows were fragmented, patterns weren’t reusable, and users were navigating across disconnected experiences to complete tasks.
SECTION
Problem
From a user perspective, this showed up as friction in core workflows like vendor onboarding, contract review, and risk assessment.
From a team perspective, it made the product hard to scale, design decisions weren’t consistent, and engineering had to constantly rebuild patterns.
before screens of Leah Legal
SECTION
Approach
I approached this from both a product and systems perspective.
First, I focused on understanding how users actually moved through the product.
For example, in a contract review workflow, the core user goal was to upload a contract and quickly understand key terms like parties, dates, liability, and payment without reading the full document.
I broke this down into task flows—from upload, to processing, to reviewing summaries and mapped both the happy path and edge cases like file errors, AI accuracy concerns, and sensitive data handling. I also designed for failure states—like when parsing fails, when data is incomplete, or when the AI might be uncertain.
From there, I started designing and refining key product experiences
SECTION
Feature-Level Work
This work led to a core AI feature that involved contract upload and summarization, where users could drag and drop a document and receive a structured, plain-language summary.
This included:
Executive summary
Key terms (termination, payment, liability)
Metadata like effective and expiration dates
AI prompt suggestions and inline refinement
Clear messaging around AI accuracy and limitations
The goal was speed, clarity, and trust helping users understand contracts in under 30 seconds.
Since this was a core AI feature, I focused on making sure information was structured and easy to scan, and designed for cases where users might question or need to double-check the output, from extractions of the data or the users own interpretation. Making sure a human remained in the loop is foundational.
SECTION
Feature-Level Work
Beyond that, I worked on core product areas like contracts, vendor onboarding, evaluation, and risk dashboards.
I restructured these workflows to:
Reduce unnecessary steps
Surface key data more clearly
Make status and progress visible
Surface AI trust signals
Improve decision-making with clearer visualizations and summaries
For example, in vendor dashboards, I redesigned how risk and compliance data was presented so users could quickly assess vendor health without digging through multiple screens.”
I also explored a chat-based interaction layer where users could interact with contracts and data conversationally, supporting onboarding, prompting, and follow-up questions.
This included things like:
Suggested prompts
Inline editing
Context-aware responses
SECTION
Feature Mapping
+ System Thinking
As I worked across these features, I mapped out how functionality and metadata connected across the system—things like summaries, chat, uploads, and data models.
This helped identify gaps and redundancies, and ensured we were building toward a more cohesive product rather than isolated features.
SECTION
Design System Auditing
Once patterns started to emerge across features, I introduced more structure through a design system.
The goal was to make these workflows scalable across multiple products and reduce rework for engineering.”
I also conducted a full audit of components and patterns across the platform to identify inconsistencies and standardize behavior.
For this, I developed a Figma plugin that uses AI to help analyze images and sort them into categories, this gave us a high-level view of the amount of UI assets and UX patterns we were using and helped lay a good foundation for discussion on implementing a design system.
SECTION
Design System Implementation
The success of auditing the live UI helped align engineering and design, and we got the greenlight to begin working on the Design System which would eventually be used across all products in the ecosystem.
Design Audit
SECTION
Refactoring
In parallel, I worked on refactoring parts of the Leah Procure experience.
In parallel, I worked on refactoring parts of the Leah Procure experience.
The original version lacked clear structure and hierarchy, so I redesigned it to:
Improve navigation and layout clarity
Surface key actions earlier
Better align the UI with user workflows
This resulted in a more intuitive and scalable foundation for future features.
Before - Middle
Final
SECTION
Collaboration + Process
A big part of the work was cross-functional alignment.
I partnered closely with product and engineering to define workflows, validate feasibility, and ensure we were solving the right problems.
I also pushed to move the team from Adobe XD to Figma, which improved collaboration, transparency, and speed across teams.
SECTION
Outcome
Overall, this work:
Simplified complex workflows across the platform
Improved how users understood and interacted with contracts and vendor data
Reduced inconsistencies across products
Created a scalable foundation for upcoming
AI-driven features
