Neuroscience PhD turned software engineer, looking for TechBio / neurotech / bio x ML roles.
My PhD was in disease modeling and mechanism in focal epilepsy. As the lab's first grad student I built its depth-electrode video-EEG system from scratch — surgery, acquisition, analysis — and wrote the Python/R signal analysis behind two papers.
I'm currently a founding partner and full-stack developer at SkyPorch, where we launched Attached, a health and wellness app on iOS and Android, and where our work on SpatialCut--a visionOS-native spatial video editor — was featured on the front page of the Vision Pro App Store and earned us an invitation to Apple HQ.
I'm also tinkering in ML: I replicated a NeurIPS 2024 result on belief-state geometry and extended it to characterize how long a small transformer holds on to predictively defunct information, published on LessWrong [<a href="https://www.lesswrong.com/posts/vzav5kfbRCDQyEB8v/toy-transformers-may-represent-belief-state-geometry" rel="nofollow">https://www.lesswrong.com/posts/vzav5kfbRCDQyEB8v/toy-transf...]
LGOICOUR · · focus · HN ↗
Remote: Yes
Willing to relocate: No (Bay Area or remote)
Technologies: Python, R, PyTorch, signal processing (EEG/LFP), React Native, Swift/visionOS, JavaScript
Résumé/CV: <a href="https://luisgoicouria.com" rel="nofollow">https://luisgoicouria.com
GitHub: <a href="https://github.com/LGOICOUR" rel="nofollow">https://github.com/LGOICOUR
Email: luisgoicouria.karate923@slmails.com
Summary:
Neuroscience PhD turned software engineer, looking for TechBio / neurotech / bio x ML roles.
My PhD was in disease modeling and mechanism in focal epilepsy. As the lab's first grad student I built its depth-electrode video-EEG system from scratch — surgery, acquisition, analysis — and wrote the Python/R signal analysis behind two papers.
I'm currently a founding partner and full-stack developer at SkyPorch, where we launched Attached, a health and wellness app on iOS and Android, and where our work on SpatialCut--a visionOS-native spatial video editor — was featured on the front page of the Vision Pro App Store and earned us an invitation to Apple HQ.
I'm also tinkering in ML: I replicated a NeurIPS 2024 result on belief-state geometry and extended it to characterize how long a small transformer holds on to predictively defunct information, published on LessWrong [<a href="https://www.lesswrong.com/posts/vzav5kfbRCDQyEB8v/toy-transformers-may-represent-belief-state-geometry" rel="nofollow">https://www.lesswrong.com/posts/vzav5kfbRCDQyEB8v/toy-transf...]