About me
> PROFILE LOADED // OK
I build AI that thinks in groups, watches closely, and learns from almost nothing. My systems borrow from swarms, hives and ghosts: agents that decide together, networks that route like colonies, models that spot what is hiding in a crowd.
My projects cover the full loop of applied AI. Seeing: FaceFit 3D, Cipher Yūrei and real-time sign language recognition. Defending: Operation Mirror Match, an adversarial fraud detection system. Deciding: Swarm Brain, Network Hive and an AI trading system that blends technical indicators, local LLMs, multi-agent analysis and evolutionary methods.
Then there is Graph Ghost, my experiment in learning with zero training through graph-based semi-supervised learning. I like the problems where the textbook answer is not enough.
Skills
Projects
Swarm Brain
A multi-agent AI framework for collaborative decision-making.
Operation Mirror Match
An adversarial fraud detection and intelligence system.
FaceFit 3D
A model-based 3D facial reconstruction and localization engine.
Network Hive
A bio-inspired intelligent routing system combining spiking neural networks and optimization algorithms.
Real-Time Sign Language Recognition
Sign language recognition using deep learning and temporal sequence modeling.
Cipher Yūrei
A computer vision platform for lip reading and visual speech recognition.
Graph Ghost
An experimental zero-training machine learning approach using graph-based semi-supervised learning.
AI-Based Trading System
Integrates technical indicators, locally hosted LLMs, multi-agent analysis, and evolutionary methods.
Contact
Want to work together or just say hi? Pick a corner of the ring.