I'm a mathematician with a background in statistics, machine learning, and data science. During my studies, I conducted research in mathematics. Today, my work focuses on full-stack development and building AI products.
I develop systems end-to-end, from application interfaces to backend services, APIs, data pipelines, and model integration.
Projects include full-stack applications for multi-agent systems, RAG pipelines, speech-to-speech interfaces, coding agents, real-time vision segmentation.
Semi-automated interview system that conducts initial screenings via voice. Saves recruiters 43 hours per month.
Self-correcting agents that generate and validate SQL queries. 68% accuracy on BIRD dataset, 17.2% above competitors.
Full-stack marketplace for AI models and APIs. Handles authentication, billing, and enterprise deployment.
Vector embedding system for comparing resumes, cover letters, and technical documents with semantic search.
AI-powered tool that converts invoice images/PDFs into structured Excel data. Saves 10 hours per 100 invoices.
Isolated environment for executing LLM-generated code with automated testing. Prevents injection attacks.
Meta-agent that continuously improves prompts for downstream systems. 18% effectiveness boost.
Lightweight application for structured note taking with custom organization system and smart tagging.
Small model trained for reliable tool-call in custom calendar management with optimized performance.
Made a tool to autmatically screen job descriptions to match experience and desired profile. Added auto-application for great fits.
Network video recorder with automated detection and event tracking. Operator-focused interface for real-time situational awareness in field operations.
Retrieval pipelines compressing input context by 10×. Multimodal RAG systems for rapid component search and UI standardization.
SQL
JavaScript
C++
gVisor
LanceDBGot a complex problem? Need a full-stack AI engineer? Let's talk.