Viewpoint eDiscovery

Viewpoint is a legal and compliance analytics platform used to manage document review for litigation, investigations, and regulatory matters. The product suite includes a modern web-based platform and two legacy on-premise applications: Review and Processing. The long-term goal is to consolidate these tools into a single, browser-based experience under Viewpoint Web.

As our team prepared to integrate AI-assisted review features into the platform, it became clear that the existing interface and user workflows needed to be re-evaluated. The legacy-inspired design introduced usability barriers and inefficiencies that risked interrupting the experience for current users once the new features were deployed.

My focus was on modernizing the UI, improving consistency across the platform, and ensuring that updates could be introduced with minimal disruption to ongoing legal workflows. I began by auditing the existing interface and any prior style documentation. From there, I developed an expanded UI guideline system and created a comprehensive component library to support future development and maintain design consistency. These standards are documented in the LCA UI Guidelines and reflected throughout the new Viewpoint Web experience.

During my initial audit of the web interface, there are two key usability and design issues that were negatively impacting user efficiency, accessibility, and overall experience: navigation and action separation, and component library and design system updates.

Actions Taken

To address the identified usability and design issues, I led a structured redesign process focused on consistency, accessibility, and efficiency:

Conducted comprehensive design audits to identify inconsistencies, accessibility failures, and areas for improvement.
Mapped user flows and created wireframes to resolve workflow friction and optimize screen layouts.
Prioritized high-frequency actions through RICE scoring, and presented my findings with strategic memos.
Developed and presented interactive prototypes to product owners, stakeholders, and lead developers for iterative feedback and alignment.

Documented all design updates within epics and supporting materials to ensure smooth handoff to development.
Performed a post-QA design audit to verify implementation accuracy and maintain design integrity across the final build.
Reorganized and simplified navigation and action menus to establish clear hierarchy and reduce cognitive load.

RICE Prioritization

The following RICE assessments illustrate how I evaluated and prioritized initiatives based on user impact, supporting research, implementation effort, and strategic value.

Navigation and Action Separation (4.3)

Reach (9/10)

Navigation affects nearly every user interaction and workflow within the application.

Impact (3.0)

Reduced cognitive load, improved discoverability, increased task completion speed, and established a scalable interaction model.

Confidence (95%)

Validated through user interviews, usability testing, click-path analysis, support trends, prototype testing, and first-click testing.

Effort (6)

Required UX research, design exploration, engineering collaboration, and implementation across shared navigation patterns.

AI-Assisted Tagging Protocol (3.4)

Reach (5/10)

Used by legal reviewers and administrators rather than the entire user base, but impacted every matter utilizing AI-assisted review.

Impact (3.0)

Improved reviewer efficiency, increased consistency of AI-generated tagging, reduced manual effort, and strengthened user confidence in AI-assisted workflows.

Confidence (90%)

Based on stakeholder requirements, workflow analysis, iterative design reviews, and prototype validation.

Effort (4)

Moderate implementation involving UX, engineering, and AI configuration without requiring significant architectural changes.

Component Library and Design System Updates (3.4)

Reach (10/10)

Every screen, workflow, and future feature relied on the shared component library and UI guidelines.

Impact (2.5)

Improved design consistency, accessibility compliance, development efficiency, and long-term maintainability across the platform.

Confidence (95%)

Supported by interface audits, accessibility reviews, engineering collaboration, and repeated implementation across multiple projects.

Effort (7)

Required extensive auditing, documentation, component creation, and cross-functional adoption.

Prioritization Summary

While all three initiatives delivered meaningful value, they addressed different strategic objectives:

Navigation and Action Separation received the highest priority because it improved the experience for nearly every user interaction while reducing friction across the platform.
AI-Assisted Tagging Protocol delivered significant value to a specialized user group by increasing efficiency, transparency, and trust in AI-assisted document review.

Component Library and Design System Updates served as foundational infrastructure, enabling consistent, accessible, and scalable product development. Even though it shares an equal score to our AI-assisted review feature, redesign requires a significant amount more effort.

Release 1: Navigation and Action Separation

Separation of Actions and Navigation

Overview

This initiative addresses a foundational usability issue within the product’s information architecture: the conflation of navigation elements and actionable controls. Based on cross-method research, including user interviews, usability testing, click-path analysis, and support ticket review, users consistently demonstrated confusion in distinguishing between “where they are going” and “what action they are taking.” This ambiguity introduced unnecessary cognitive load, increased decision friction, and contributed to navigation errors and task delays, particularly for new and infrequent users.

Problem Statement

The current interface blends navigational links and functional actions within the same UI regions, requiring users to pause and interpret intent before each interaction. This mismatch between system design and user mental models results in:

Increased hesitation before clicks (decision friction).
Misinterpretation of interactive elements.

Slower task completion and inefficient workflows.
Reduced confidence in navigation, especially for new users.

Strategic Rationale

Navigation systems should communicate location, while UI actions should communicate intent. The lack of separation between these two constructs creates ambiguity that scales poorly as product complexity increases. Establishing a clear separation between navigation and actions introduces a more predictable interaction model, improving both usability and long-term maintainability of the design system.

Proposed Framework

This separation ensures that users can reliably distinguish between “moving through the system” and “performing work within it.”

Navigation Layer (System Orientation)

Side Drawer: Primary wayfinding and destination structure.
Breadcrumbs: Contextual location awareness.
Top Ribbon (Navigation Context Only): Historical or hierarchical location cues.

Action Layer (Task Execution)

Dedicated Action Zones separated from navigational structures.
Top Ribbon (Action Context Only): Task-based triggers and state changes.
Grouped by priority and frequency of use.

Validation Approach

Findings indicated improved clarity, reduced hesitation, and more direct task completion paths. The framework was validated through iterative prototyping and testing:

Prototype usability testing to assess comprehension of structure.
First-click testing to measure interaction accuracy.

Task success rate analysis to evaluate efficiency gains.
Comparative evaluation of pre/post navigation models.

Outcome Goals

Findings indicated improved clarity, reduced hesitation, and more direct task completion paths. The framework was validated through iterative prototyping and testing:

Reduce cognitive load during navigation.
Increase first-click accuracy.
Improve task completion speed.

Decrease navigation-related errors.
Establish a scalable interaction model for future product growth.

Summary

Separating navigation from actions resolves a core structural ambiguity in the product experience. By aligning interface behavior with user mental models, this initiative improves usability, reduces friction, and creates a more scalable foundation for future feature expansion.

Release 2: AI-Assisted Tagging Protocol

Enhanced Review Setup

Documents Review

Enhanced Review Setup

Overview

As part of the Viewpoint eDiscovery redesign, I contributed to the design of an AI-powered tagging feature that assists legal reviewers by automatically identifying documents that may require specific review classifications. Rather than replacing reviewer judgment, the feature augments existing workflows by surfacing AI-generated tag recommendations based on contextual patterns, case criteria, and document content.

The objective was to reduce repetitive manual effort, improve consistency across review teams, and increase confidence in AI-assisted decision-making through transparent, explainable recommendations.

Problem Statement

Legal document review workflows require reviewers to evaluate large volumes of content against complex and evolving criteria, including relevance, privilege, confidentiality, and case-specific rules. These decisions often rely on manual interpretation of keywords, date ranges, and contextual relationships across documents. This process presented several issues:

Inconsistent tagging across reviewers and teams.
High cognitive load due to complex rule interpretation.
Time-intensive manual classification of documents.

Difficulty scaling consistent review standards across large matters.
Limited transparency in how tagging decisions were derived.

Strategic Rationale

As Viewpoint advanced toward integrating AI into core review workflows, it was critical that automation enhanced, not obscured, reviewer decision-making. In legal and compliance contexts, trust, transparency, and auditability are essential requirements, particularly when AI influences document classification. This feature was prioritized because it directly supports the responsible adoption of AI by ensuring that recommendations are explainable, configurable, and aligned with established legal review standards. By structuring tagging logic into clear, human-readable components, the system reduces friction while preserving reviewer authority and defensibility of decisions. From a product perspective, this initiative also establishes a scalable foundation for future AI-assisted capabilities by standardizing how AI reasoning is represented in the interface.

Proposed Framework

This structure was designed to make AI behavior interpretable, configurable, and aligned with legal review workflows while maintaining reviewer authority over final decisions. To address these challenges, I assisted to design an AI-assisted tagging protocol structured around five key components:

Overview

Defines the purpose of the tagging protocol, including the classification intent (e.g., relevance, privilege, confidentiality) and how it applies within the review workflow.

Dates

Establishes relevant time ranges associated with case or matter context to support temporal filtering and interpretation.

Keywords and Phrases

Identifies critical terms, entities, and language patterns used to guide AI tagging logic and improve contextual matching.

Criteria

Defines explicit logical rules that combine inputs such as keywords and date ranges (e.g., documents containing both [keyword A] and [date range B] are flagged for review).

Notes

Provides space for reviewer or administrator guidance, clarifications, and exceptions to support consistent interpretation and decision-making.

Validation Approach

Validation focused on clarity of logic presentation, reviewer trust in AI suggestions, and the ability to correctly interpret tagging rationale without additional training. The framework was validated through a combination of qualitative and iterative evaluation methods:

Stakeholder review sessions with legal and compliance teams.
Workflow analysis of existing tagging behaviors and decision patterns.
Prototype testing with representative review scenarios.

Iterative refinement based on usability feedback and edge-case testing.
Comparative evaluation of manual vs. AI-assisted tagging workflows.

Outcome Goals

The enhanced tagging framework was designed to achieve the following outcomes:

Reduce manual effort required for document classification.
Improve consistency and accuracy of tagging across reviewers.
Increase transparency in AI-generated recommendations.

Accelerate review workflows through contextual suggestions.
Support scalable application of complex review criteria
Improve reviewer confidence in AI-assisted decision-making.

Summary

The Enhanced Review AI feature introduces a structured, explainable approach to AI-assisted tagging within legal document review workflows. By organizing tagging logic into clear, interpretable components, the framework improves transparency, reduces cognitive burden, and enhances consistency across review teams. This approach ensures that AI functions as a supporting layer rather than a replacement for reviewer judgment, enabling faster, more reliable review processes while maintaining strict user control and interpretability.

Release 3: Component Library and Design System Updates

Homepage

Document Review

Homepage, Old UI

Dashboard, Old UI

Document Review, Old UI

1/1

Overview

As part of the Viewpoint Web modernization initiative, I led efforts to redesign core user interface patterns and expand the existing component library to support a more consistent, accessible, and scalable product experience. The platform had evolved over time through incremental updates and legacy constraints, resulting in visual and structural inconsistencies across key workflows. The objective of this initiative was to modernize the UI while establishing a unified design system that could support both current functionality and future feature expansion.

Problem Statement

Visual issues created friction for users navigating the platform, and slowed development cycles as teams repeatedly rebuilt similar UI elements without a shared system of record. The existing interface and supporting design assets presented several challenges that impacted usability, development efficiency, and long-term scalability:

Inconsistent UI patterns across screens and workflows.
Legacy design elements that did not align with modern usability standards.
Fragmented component usage with duplicated or conflicting patterns.

Accessibility gaps impacting WCAG/ADA compliance.
Lack of standardized documentation for UI behavior and interaction rules.
Increased engineering overhead due to repeated or custom-built components.

Strategic Rationale

A unified component library and updated UI system were necessary to ensure consistency, reduce user friction, and enable scalable product development. Without a standardized system, each new feature risked introducing additional inconsistencies, further compounding usability and maintenance challenges over time. This initiative was prioritized because it addressed a foundational layer of the product experience. Unlike feature-level improvements, design system updates have a compounding impact across all workflows, improving usability, accessibility, and development velocity simultaneously. Additionally, establishing a cohesive UI system was critical to supporting future enhancements, which require predictable interaction patterns and accessible design standards.

Proposed Framework

Together, these systems established a single source of truth for UI behavior and visual standards across Viewpoint Web. The modernization effort was structured around two interconnected components:

UI Redesign System

Standardized layout structures across core workflows.
Updated visual hierarchy to improve readability and reduce cognitive load.
Consistent application of typography, spacing, and color usage.
Alignment with modern accessibility and usability standards.
Reduction of visual clutter and redundant interface elements.

Component Library Expansion

Creation of reusable, documented UI components.
Standardized interaction patterns for common actions (e.g., modals, tables, filters).
Defined states, behaviors, and usage guidelines for each component.
Centralized design system documentation (LCA UI Guidelines).
Alignment between design and engineering implementation through shared specifications.

Validation Approach

Validation focused on usability improvements, consistency of interaction patterns, accessibility compliance, and ease of implementation within engineering workflows. The redesigned UI system and component library were validated through a combination of iterative and cross-functional evaluation methods:

Heuristic usability reviews to identify friction points in existing workflows.
Accessibility audits aligned with WCAG/ADA standards.
Cross-functional design reviews with engineering and product stakeholders.

Prototype testing of redesigned components in real workflow scenarios.
Comparative evaluation of legacy vs. redesigned interface patterns.
Iterative refinement based on implementation feedback during development cycles.

Outcome Goals

The UI redesign and component library update were designed to achieve the following outcomes:

Improve consistency across all user interfaces and workflows.
Reduce cognitive load through clearer visual hierarchy and interaction patterns.
Increase accessibility compliance across the platform.

Improve development efficiency through reusable components.
Reduce UI duplication and design inconsistency across teams.
Establish a scalable foundation for future product features and AI integration.

Summary

The UI redesign and component library initiative established a unified design system for Viewpoint Web, addressing long-standing inconsistencies while laying the foundation for future scalability. By aligning visual design, interaction patterns, and component behavior into a single system, the initiative improved usability for end users and efficiency for development teams. This work transformed the UI from a collection of legacy-driven patterns into a structured, maintainable design system capable of supporting ongoing platform evolution and enterprise-scale feature development.