I build LLM assistants, production APIs, and developer tools — from retrieval and tool calling to backend architecture, observability, and client applications.
|
A shared-expense ecosystem with a SwiftUI client, Python backend, and Telegram Mini App. Developed as a project with a mentor-engineer from Yandex.
|
An admissions workflow platform for Sirius with role-based access, guarded data imports, reviews, interviews, analytics, and AI-assisted checks.
|
|
A self-hosted Telegram personal assistant. I am developing multi-user isolation, secure Telegram onboarding, resumable contact-graph analysis, bounded model calls, and safer owner routing.
|
|
| Area | Technologies and practices |
|---|---|
| AI / ML | LLM integrations, assistants and agents, tool calling, RAG, guardrails, structured output, local inference, Computer Vision |
| Backend & data | Python, FastAPI, SQLAlchemy, REST / OpenAPI, PostgreSQL, MongoDB, Kafka, NumPy, pandas |
| Infrastructure | Docker, GitHub Actions, S3, Prometheus, Grafana, Loki, integration and E2E testing |
| Clients | TypeScript, React, SwiftUI, Telegram Mini Apps |
- Sirius University of Science and Technology — Experimental IT Specialist program
- AI Agents Security Week — Yandex School of Data Analysis (SHAD) and Yandex
- Machine Learning & Linear Algebra — Sirius, completed with honors
- Ethical Hacking with Python — 148-hour course
AI Engineer · Python Backend · Moscow / Sirius