AI is my build system. I learn fast, dive deep, and move from idea to working product — shipping tools, platforms, and intelligence into the real world.
AI-native products conceived, built, and pushed into the real world — not stranded in a deck.
A social platform built from zero to 20K+ users, with fast iteration and AI-assisted product development at its core.
A verified, campus-only social layer where official events and spontaneous plans finally meet.
Founder outcomes, AI product velocity, and research impact — one operating range.
Four frontiers. One method: find the signal hiding inside the complexity.
From first research models to products in market — a deliberately non-linear path.
Heart murmur classification · Trinity College
100K+ weekly AI requests · Radical AI, NYC
Yaaro launches · LIGO research · Quantum control
ConnectED live · Incoming AI SWE at Zemantics · Trinity graduate
Ghost Vectors: Soft-Deleted Embeddings Remain Reconstructible in HNSW Vector DatabasesARXIV · AI SECURITY
↗ 02Deep Reinforcement Learning for Autonomous Control of Hole-Doped Hubbard ClustersPUBLISHED · APL MACHINE LEARNING
↗ 03Microseismic Noise Mitigation with Machine Learning for Advanced LIGOARXIV · GRAVITATIONAL WAVES
↗ 04Comparative Analysis of Autonomous and Systematic Control Strategies for Hole-Doped Hubbard ClustersARXIV · REINFORCEMENT LEARNING
↗ 05Geometric and Orbital Control of Correlated States in Small Hubbard ClustersARXIV · QUANTUM SYSTEMS
↗ 06Prediction of Neutrino Emissivity from Pre-supernova Stars by Machine LearningPUBLISHED ABSTRACT · APS
↗