Legal Discovery Request GAI Workflow Tool
The regulatory legal team at Duke Energy needed a more consistent, efficient way to manage over 20,000 discovery requests received annually across jurisdictions. Existing processes and tools were fragmented, leading to inefficiencies, duplicated effort, and inconsistent tone in responses. Before defining any solution, I led exploratory research to understand user pain points, decision flows, and areas where AI could provide measurable impact.
My Role
I began as a facilitator for the initial design thinking workshop and, once the product was funded, continued as an embedded strategist with 50% allocation to the product team. My responsibilities included framing the problem space through research, aligning cross-functional stakeholders using design thinking methods, and introducing measurable practices that informed decision-making in the absence of a clearly defined MVP.
​
I also served as a stabilizing presence during periods of turnover, mentoring and supporting the product owner, business analyst, UX designer, and change manager. My focus was not on UX design, but on research, facilitation, and the creation of alignment frameworks that enabled the team to move forward cohesively in a complex and evolving environment.
Facilitation & Alignment
Facilitated cross-functional workshops to align on current-state pain points and define future-state capabilities.
UserZoom survey established baseline metrics for consistency and speed in discovery requests.
Decision Frameworks
Structured intake and prioritization discussions using desirability, viability, and feasibility to balance perspectives.
UserZoom survey established baseline metrics for consistency and speed in discovery requests.
Vision Workshop
Facilitated a product vision workshop to align the team around a shared direction. Even with limited input, I synthesized discussion into a vision statement and a set of experience principles to guide product decisions going forward.
UserZoom survey established baseline metrics for consistency and speed in discovery requests.
Approach & Strategy
Building on the research findings, I led collaborative workshops and introduced strategic frameworks that translated insights into clear direction.
Research & Insights
To understand how discovery requests were handled and where generative AI could create measurable impact, I conducted mixed-method research - combining interviews, synthesis, and quantitative validation.
Interview findings revealed consistent pain points across role types and workflows.
01 | User Interviews
-
Conducted 21 interviews across SME, paralegal, and attorney roles to uncover workflow pain points.
-
Identified recurring breakdowns in version control, request duplication, and lack of traceability.
02 | Synthesis in Dovetail
-
Tagged and themed 250+ data points in Dovetail, clustering insights by user role and workflow stage.
-
Translated findings into key opportunity areas visualized in journey maps
Synthesis in Dovetail revealed three key opportunity themes: visibility, consistency, and efficiency.
03 | Survey Validation (UserZoom)
-
Designed and deployed a baseline survey to quantify user frustration and time loss.
-
Connected qualitative insights to measurable performance metrics to track improvement potential.
UserZoom survey established baseline metrics for consistency and speed in discovery requests.
Impact
Aligned stakeholders around a data-informed understanding of user pain points, shifting decision-making from assumption-based to evidence-based.
Delivered a product vision statement and experience principles that gave the team a guiding north star in the absence of clear MVP direction.
Introduced structured research insights in Dovetail, creating a traceable link from user pain points to product opportunities.
Designed and deployed the first-ever measurement framework via a UserZoom survey, providing baseline data for evaluating future MVP success.
Enabled leadership and stakeholders to make more informed decisions by translating ambiguity into decision-ready artifacts.
Reflection
This project highlighted the value of design strategy in ambiguous product environments. While the MVP itself remained undefined, my work created the infrastructure of clarity: facilitating workshops, synthesizing user research, and introducing measurement where none had existed. These contributions not only advanced the conversation around AI integration in legal discovery but also demonstrated how design strategy can bring alignment, structure, and user-centered direction to complex, evolving product spaces.