applied AIAI adoptionresearch

work & research.

a selected view of the systems, publications, and projects behind my profile — AI products shipped under startup constraints, research with measurable outputs, and tools that make AI more accessible.

reach out

publications

research artifacts.

PMLR 2025

Towards an efficient, customizable, and accessible AI tutor

an offline RAG pipeline paired with a lightweight language model to support educational tutoring in low-resource settings.

Modular memorability

a modular computer vision model for video memorability prediction — connecting visual understanding with cognitive performance signals.

ML-SPEAK

a theory-guided machine learning method for predicting conversational turn-taking patterns from team composition.

where the work met reality.

ML engineer · head of data management

Memorable

built and maintained computer vision models for advertising optimization, then designed the data infrastructure and processing pipeline that enabled 10× storage growth. Memorable was acquired by Reddit in 2024.

external reference↗

Rice MS data science capstone

NASA navigation assistance

built a two-stage crater detection system for spacecraft localization with record precision and presented the work to NASA engineers in Houston.

— internal work

Universidad Austral

AI, innovation & government

coordinate academic work and applied programming that helps public-sector professionals connect AI foundations with governance, policy, and institutional use cases.

— internal work

selected GitHub work