Work

Research before recommendations.

Observe

Watch how people actually work

Evidence

Data plus context, never one alone

Verdict

A decision someone can act on

The investigations

Investigation 01

High-level · NDA-safe

When trust determines adoption.

UX research · Eli Lilly, via Moral Agency

The situation: Inside a Fortune 500 pharmaceutical company, AI and automation tools were arriving faster than people could learn to trust them — and the people expected to use them have expensive time and long memories.
What I noticed: The usability protocols measured what the team hoped to see, not how people actually worked. Priorities lived in meetings and never made it into requirements. A tool nobody trusts doesn't get a second chance.
The call: Rebuild the protocols around real behavior — study design, moderated sessions, synthesis — and turn stakeholder workshops into requirements a team could actually build against.
What followed: ▸ The requirements were adopted across concurrent initiatives, and the release they shaped tested comfortably above the industry usability benchmark — with the users whose trust is hardest to win.

Investigation 02

Analytics · SAP

The situation: Marketing and sales tracked their performance across scattered reports that disagreed with each other.
What I noticed: Teams spent more time finding data than using it — and while they searched, competitor moves and open space in the market went unexamined.
The call: Build one library. Standardize how performance is tracked across both organizations, and put competitor and market analysis beside it, where decisions get made.
What followed: ▸ It became the single source of truth — marketing aimed sharper, and sales stopped hunting for data.

Investigation 03

UX research · BDPA.org

The situation: A live site serving a national technology community — approached as a client engagement, with real people mid-task on it every day.
What I noticed: Watching the site perform under real use — moderated sessions, stakeholder interviews, the information architecture under load — told a different story than the sitemap did.
The call: Put the research in front of the redesign: personas, journey maps, and Figma recommendations built from observed behavior, not taste.
What followed: ▸ The client adopted the research as the blueprint for the redesign.

Skills

SQL · Python · Tableau · Power BI · Excel (Microsoft Certified Expert) · Data visualization

UX research (Professional Certificate) · Usability testing · Journey mapping · Figma · Product strategy

Agile / Scrum · Jira · Confluence · Requirements gathering · Process improvement · Stakeholder management

Generative AI · Prompt engineering · AI for design & optimization · SAP ecosystem · Enterprise systems (ERP)

Additional projects

AI-assisted product build, concept to high-fidelity Figma prototype — requirements gathering, Agile delivery with Jira, and user research throughout.