cs honors & neuroscience @ ut austin
i build AI that reads, models, and repairs the brain
reteena • scholar • linkedin • email
- research fellow & venture scout @ contrary
- co-founder @ reteena — applied AI lab for neurodegenerative diagnostics: remembrance (AI reminiscence-therapy companion for alzheimer's patients), TRIDENT (an adaptive brain-computer interface), lark (stakeholder-aware multi-agent decision making), geneattentionnet (interpretable gene-expression classification)
- founding @ xyla engineering — deployable biofactories for maritime environments
- software engineer intern @ nvidia — holoscan team
- software engineering intern @ pure storage — EHR latency prediction & infrastructure telemetry modeling
- researcher @ harvard medical school + mass general — ophthalmic foundation models & AI-guided surgical systems
- research intern @ memorial sloan kettering — stage-specific molecular network rewiring & graph representation learning
- software developer & researcher @ princeton — drosophila connectome mapping & computational neural circuit analysis
- researcher @ columbia (bme/neuro) — neural encoding of visual information & working memory dynamics
- cohort 1.0 @ fr8 · f.inc off season · the residency
| project | what it is |
|---|---|
| neuroforge | browser-native CAD for neural circuits — draw a circuit, watch it spike. five membrane models, real synaptic kinetics, GPU rendering. live demo |
| tribe-explorer | visual analytics for the latent spaces of neural foundation models — latent geometry, neural timelines, cross-subject alignment |
| neural-morphogen-operators | learning continuous reaction-diffusion dynamics from spatial transcriptomics |
| gauge-neural-dynamics | do biological neural representations use context-dependent coordinate systems? trying to find out |
- Lark: Biologically Inspired Neuroevolution for Multi-Stakeholder LLM Agents — NeurIPS workshop, 2025 · arXiv:2510.16978
- NeuroLoom: A Framework for Reproducible Cortical Microcircuit Modeling with Spiking Neural Networks — ICLR FM4Science workshop · code
- the rest are on google scholar