LLM features that survive production — retrieval, tool calling, structured outputs — shipped end to end on React and FastAPI

Open to worldwide remote AI product engineering roles

Languages: Russian — Native · English — IELTS Academic 7.5 (C1)

Email · Telegram · LinkedIn · GitHub · Website

• AI Reliability: Structured outputs with schema validation, tool calling that handles partial failure, async orchestration and streaming that survives reconnects
• Full Product Surface: I ship the React interface and the FastAPI service behind it, so AI features do not stall on a backend handoff
• Legacy Frontend Migration: Angular and Svelte to React in production, without freezing feature delivery

At T-Bank I ship product work on an internal LLM assistant used by 38,000+ employees, moving already built logic from Svelte to React while feature delivery keeps running, and I own the FastAPI services behind the interfaces I build

Internal web application for an AI image-generation platform used by designers who needed a practical interface for day-to-day creative work. The product provided board-based organization, editing workflows, and async task management on top of a company-wide generation API

Reach: 1,000+ internal users.

• Built and evolved a board-based workspace where designers organized generations into project spaces instead of a flat task list
• Implemented drag-and-drop between boards, image reactions, and editing flows for multiple generation modes
• Designed a priority-queue polling model on the frontend to schedule status checks for asynchronous generation tasks without aggressive polling

Company-wide internal LLM assistant used by 38,000+ employees. The product is built on a mixed Svelte and React frontend with a FastAPI backend and is progressively moving toward a React-first architecture without pausing feature delivery

Reach: 38,000+ employees.

• Worked in a mixed frontend stack, helping move product logic into React-owned areas while maintaining the existing Svelte surface
• Shipped product features and supporting backend fixes in parallel with migration work instead of treating them as separate tracks
• Helped establish stricter frontend quality rails around new React code while the product still carried legacy surface area

• StudHack — Platform for hackathon teams, events, and career communities with user profiles, team workflows, notifications, and Dockerized UI/API delivery
• FastAPI MVP — A reusable FastAPI bootstrap package with Mongo and S3 helpers, metrics, and stricter environment management for new service setup
• OpenAI Async Functions — Async Python library for OpenAI-compatible tool calling, with Pydantic request and response models, runtime tool registration, and structured output helpers
• lint-gost-tex — CLI linter for LaTeX documents against GOST-style formatting rules, with checks for references, lists, captions, forbidden abbreviations, Unicode, and RU/EN spelling

If you are hiring for AI product engineering, LLM reliability work, or React and FastAPI product delivery, the fastest way to reach me is by email, Telegram, or LinkedIn.