Computer Science student at Saint Louis University (class of 2027), originally from Tbilisi, Georgia. I mostly build backends, data pipelines, and ML / computer-vision tools.
- Undergraduate research assistant in SLU's Networking Lab, surveying eBPF for programmable networking
- Co-founder and developer at SkyInsights
- Most recent project was a warehouse staffing platform for a logistics client
J.U.T.S.U. — Joint Upper-limb Tracking System for UI augmentation
Real-time hand gesture recognition that renders GPU-accelerated effects onto a live virtual camera in Discord and OBS. Custom scikit-learn ensemble over an 83-dimensional hand pose feature vector, with zero CPU/GPU usage at idle.
Python MediaPipe scikit-learn CuPy OpenCV
BachmannLogi — Warehouse staffing management system (private client project)
Replaces an all-Excel workflow: async FastAPI/SQLAlchemy backend, React + TanStack Query frontend, and Excel ingestion pipelines feeding an Alembic-managed PostgreSQL schema. EU-only hosting and JWT verification against Supabase JWKS for GDPR-regulated data.
FastAPI React TypeScript PostgreSQL Supabase
Vigilane — AI road hazard detection (HackSLU 2026, team of 4)
Owned the database and API layer of a real-time dashcam hazard detection system: Firestore service modules, REST blueprints consuming a YOLOv8 inference pipeline, and Fernet encryption for sensitive fields at rest.
Python Flask Firestore YOLOv8
Binge — Recipe recommendation web app
Designed the PostgreSQL schema and REST endpoints, a daily GitHub Actions pipeline seeding recipes from the Spoonacular API, and a frequency-weighted recommendation algorithm.
TypeScript PostgreSQL Supabase GitHub Actions
Smaller things in C++: raytracing_cpp · Process-Monitor · cpp-drills
Languages: Python, C++, TypeScript / JavaScript, SQL, Lua
Backend: FastAPI, Flask, Node.js, Express, PostgreSQL, Supabase
Frontend: React, TailwindCSS
ML & CV: scikit-learn, MediaPipe, OpenCV, CuPy, NumPy, Pandas
Tools: Git, Docker, Linux, Bash, GitHub Actions