Methodology
Research Design
The core of my practice. I design studies from the demands of the research question and constraints, not from a standard methods toolkit. This includes adversarial evaluation protocols, phased prioritization exercises that manage cognitive load while preserving quantitative and qualitative depth simultaneously, bilingual and culturally adapted stimulus materials, and multi-phase iterative study architectures that incorporate organizational buy-in as a structural feature of the research design.
See the adversarial protocol in practice →Methodology
Qualitative Methods
I match qualitative methods to research questions, not the other way around. Generative interviews when the problem space is undefined. Concept evaluation with storyboards when ideas need user input before engineering investment. RITE when the team needs to diagnose and iterate in days, not months. Focus groups when participant-to-participant dialogue will surface dynamics that individual interviews cannot. Field research and in-home studies when context shapes the finding.
See generative interviews and focus groups in practice →Methodology
Quantitative & Scaled Methods
I design quantitative frameworks that produce defensible evidence for high-stakes decisions. For the EMG evaluation, this meant developing a priori hypotheses, counterbalancing experimental conditions across 60+ participants, and computing F1 scores from paired telemetry and human-coded ground truth. At GetYourGuide, it meant building NLP-powered dashboards that processed 2.5 million customer reviews into actionable signal across design, product, and the C-suite. The common thread: choosing the right metric for the question, not defaulting to whatever the analytics platform provides.
See diagnostic quantitative methods in practice →Practice
Research Infrastructure
I have built research functions from scratch twice, at GetYourGuide and Choco. This means intake processes, insights repositories, consultation models, stakeholder engagement frameworks, and training programs that extend research capability beyond the research team. It also means Research Operations: recruiting pipelines, vendor management, and workflow systems that let a small team operate at scale.
See practice building at scale →Expertise
AI & Emerging Technology
Four years at Meta Reality Labs Research across product spaces that did not yet have established benchmarks: agentic systems, always-on AI assistants, contextual AI, and multimodal interaction including gestural input, EMG, and camera vision. This work required designing studies where standard methods would produce misleading results, and generating evidence credible to both engineering and product leadership.
See longitudinal AI device evaluation in practice →Expertise
International Research
Cross-cultural study design, culturally adapted stimulus materials, interpreter-mediated research protocols, and analytical frameworks grounded in cultural and historical context. Field work conducted across six countries on three continents.
See cross-cultural research in practice →