Software developer focused on AI-powered products, reliable backend systems, and practical product engineering.
I build end-to-end products: from domain models and persistence to external integrations, tests, release automation, and native interfaces. My current work spans Python/LLM systems and SwiftUI applications.
Local-first persistent AI character core with explicit state ownership, layered memory, and replaceable LLM providers. Built with Python, SQLAlchemy/Alembic, Ollama/OpenAI integrations, embeddings, strict typing, and an extensive test suite.
Resilient Telegram job-search agent with HeadHunter OAuth, structured LLM ranking, provider abstraction, retries, and persistent workflows. Built with async Python, SQLAlchemy/Alembic, PostgreSQL, Docker, and automated tests.
Local-first iOS learning app built around finite sessions, deterministic ranking, spaced recall, and a persistent Growth Memory. Built with SwiftUI, SwiftData, and Swift Testing.
Native recipe companion with authentication, asynchronous API integration, local favorites and history, localization, WidgetKit, accessibility work, and release verification.
Telegram product for EGE/OGE mathematics preparation with student quizzes, teacher tools, homework workflows, learning materials, and progress tracking.
Python 路 asyncio 路 Pydantic 路 SQLAlchemy 路 Alembic 路 PostgreSQL 路 SQLite 路 LLM integrations 路 Ollama 路 OpenAI 路 Swift 路 SwiftUI 路 SwiftData 路 Docker 路 GitHub Actions 路 pytest 路 Swift Testing
AI/LLM systems with controlled state, backend reliability, and product engineering across Python and native iOS.