professional experience
I translated research findings into journey maps, challenge frameworks, and adoption strategies across public finance, enterprise data, and AI-enabled forecasting.
Public Financial Management Advisory
Role
Service designer (synthesis & framework design)
Timeline
1 month
Methods
Journey mapping, affinity synthesis, workshops
Partnered with a national government's budget authority to redesign its budget preparation process ahead of a digital transformation initiative — ensuring new technology would improve, not just digitize, existing ways of working.
My contribution
Synthesized 50 pain points from cross-agency workshops into a challenge framework : layering our team's public finance digitalization pillars against the phases of the As-Is budget cycle, which became the shared reference for stakeholder alignment on future-state design.
Approach
The government didn't have an existing taxonomy to sort 50 pain points into. I organized them around a public finance framework as the structural backbone and layered the As-Is budget cycle on top of it as a second axis. Each pain point sat at the intersection of “which capability is breaking” and “where in the process it breaks.”
→ Framework carried into a second mission and a global workshop (100+ participants, 4.7/5 rating)
Lessons learned — regulated industries + complex systems
In a multi-agency system, front-stage pain looks completely different depending on the user — a national budget officer and a regional program office hit different walls. Mapping those journeys onto one shared pillar structure let me trace unrelated-looking complaints back to the same root cause in the underlying data and systems, and turned the framework into something stakeholders could actually align around, not just a list of findings.
Internal System UX Research
Role
UX researcher & service designer
Timeline
2 research sprints, 6 weeks
Methods
Interviews, usability testing, journey mapping
After a platform migration, staff struggled to discover internal data resources. What started as a UI audit was reframed into an institution-wide discoverability challenge.
My contribution
Designed and led interview and usability protocols; mapped a six-stage user journey (awareness → enter → navigate → interpret → exit → extend) revealing why users bypassed the official platform for informal workarounds.
Approach
Interviews kept surfacing tool names — trading terminals, statistical software, direct database queries — before we'd even asked about the official platform. I treated that as signal, not noise: the platform wasn't losing to a bad interface, it was losing to an entire data ecosystem users already trusted more.
→ Recommendations across UX, metadata, search and AI-readiness adopted into an implementation roadmap
Lessons learned — finance data tools
This project sharpened how I read financial data workflows, recognizing when low adoption of an internal tool actually means that specialized platforms already meet the need better, and designing around that reality rather than against it.
International Monetary Fund — Economic Research Initiative
Role
UX researcher
Timeline
3 months
Methods
Discovery sessions, dashboard walkthroughs
An AI-powered forecasting tool for economists in low-data environments was technically ready, but adoption was unproven — the client needed to know whether users would trust and use it.
My contribution
Ran discovery sessions and dashboard walkthroughs with economists and technical experts; synthesized findings into adoption factors, separating genuine skepticism about data quality from acceptance of the underlying concept.
Approach
The technical solution already existed by the time we joined, so the real risk wasn't feasibility — it was trust. I shifted discovery sessions away from the dashboard interface and toward how economists currently justify their forecasts, which is what surfaced transparency as the actual adoption blocker.
→ Transparency and workflow-integration recommendations shaped the platform's onboarding strategy
Lessons learned — workflow analysis
I came away better at separating “people don't like this” from “people don't trust this yet” — two signals that look identical in early feedback but call for completely different fixes.
