Projects

I study how digital and AI systems shape the lives of immigrant communities and nonprofits — often in ways that were never intended, and rarely examined.

Flagship Project
2023–Present

AMINA

PDF

An AI assistant designed with — not just for — immigrant-led nonprofit organizations, built through participatory co-design. Accountability is a feature, not an afterthought.

AMINA

AMINA addresses a specific problem: immigrant-led nonprofits are systematically underrepresented by the platforms meant to connect them with funders, because those platforms assume a kind of organization they aren't — formally incorporated, English-speaking, with a strong digital presence.

AMINA was built through 20+ interviews and four participatory co-design sessions. A central design principle: organizations can review and correct how the AI describes them. Accountability is built into the interaction, not added on top.

  • Multilingual support designed in from the start
  • Uncertainty made visible — the system flags when it's guessing
  • Community members as co-designers, not just testers
PhD Research
2021–Present

AI Infrastructures & Immigrant Nonprofits

PDF

How do donor platforms, grant databases, and AI-powered intake tools shape which organizations get seen and supported — and which don't?

Research

My dissertation examines how AI systems — donor platforms, grant databases, intake tools — shape the visibility and viability of immigrant-led nonprofits and informal community organizations. These tools are typically built for well-resourced, formally incorporated, English-speaking organizations. Groups that operate outside those norms are made less visible, less fundable, and less supported.

I use qualitative interviews, participatory design, and co-design methods. My fieldwork is based in Toronto, working with immigrant-led organizations across a range of communities.

Earlier Work
2017–2021

Brain-Computer Interfaces & Neural Data

PDF

How meditation and mental rehearsal affect EEG-based brain-computer interface classification accuracy — and what it taught me about designing for people, not averages.

BCI

My MSc at Holland Bloorview (supervised by Dr. Tom Chau) examined how meditation and mental rehearsal affect EEG oscillations and brain-computer interface classification accuracy — an early encounter with the gap between what a system is designed for and who can actually use it.

Before that, I built a data pipeline at the Ontario Brain Institute that migrated a large MRI imaging database to the Brain Imaging Data Structure (BIDS) standard, and contributed to multivariate statistical tools for outlier detection in complex neuroimaging data.

That technical foundation shapes how I work now — I can engage substantively with AI system design, not just critique it from the outside.