I study why performance degrades under pressure: whether the cost shows up in the quality of sensory evidence, in the threshold for acting on it, or elsewhere in the chain. My research uses psychophysics and computational modeling to pull apart dual-task costs in aging, where physical effort, arousal, and perception interact in ways that matter for applied human factors.
The central tool is a handgrip dual-task paradigm that combines concurrent isometric grip, auditory and visual discrimination tasks, pupillometry, and hierarchical Bayesian drift-diffusion modeling. Rather than reporting a single performance score, I model where in the decision process costs appear. That matters if you want to design systems that respond to cognitive state, not just average accuracy.
Fluent in English & Persian | Modest conversational French
Education
Neurotree profile (academic genealogy)
PhD in Psychology and Cognitive Neuroscience
University of California, Riverside
2021–2026 (Expected)
M.Sc. in Psychology
Queen’s University
2017–2020
Hon. B.Sc. in Biological Sciences & Psychology
York University
2013–2017
Research & Industry Experience
Research Scientist
thebridge Inc.
San Francisco, CA | Sep 2026 – Present
- Designing measurement frameworks for scoring the quality of group deliberation from conversation transcripts: construct definitions, behaviorally anchored rating scales, and explicit rules for when an item cannot be assessed
- Planning validation of human raters and LLM judges: reliability on ordinal ratings, human versus model agreement, and tests of whether an automated judge can stand in for a human annotator
- Treating preprocessing as a source of measurement error: how splitting a conversation into units changes scores, tested with sensitivity analyses rather than assumed away
- Building composite metrics with stated value choices and robustness checks across reasonable alternative specifications
- Writing implementation specs and automated sanity checks so evaluation pipelines can be audited
Research Assistant I
Surgical Safety Technologies
Toronto, ON | 2021
- Assisted on instance segmentation pipelines for surgical tool detection and tracking
- Computed inter-rater reliability (κ and ICC) for ML annotation quality across analyst teams
- Reviewed ML-for-surgical-safety literature to support internal research operations
- Learned that alert fatigue and interface friction, not raw model accuracy, are what sink adoption of ML tools in clinical settings
XR Interaction Research: Independent Study
Self-directed Research
2025–2026
- Built a fully remote web-based interaction testbed using React and TypeScript
- Conducted a within-subject hand vs. gaze pointing study with ISO 9241-9 / Fitts’ law design
- Published findings as arXiv preprint (arXiv:2603.15991)
- Applied hierarchical Bayesian LBA in PyMC for verification-phase response time modeling
Surgeon Cognitive Dashboard: R&D Project
Self-directed Research
2025
- Designed and deployed a real-time R Shiny dashboard for surgical trainee cognitive monitoring
- Integrated XGBoost multiclass classifier with Platt scaling for cognitive state prediction
- Built three instructional threshold policies for training contexts (Adaptive Gain, SDT, Fatigue-Adaptive)
- Live demo at: shinyapps.io