00Python developer · backend, front end, infrastructure, AI integrations
AI consultants, bots and automation for small and mid-size business — before → after.
Chatwoot + an AI bot that prices products and hands over to a manager when needed.
Starting point: ~30% of emails were answered within an hour.
AI assistant for the team in Telegram: knowledge base, Metrika, Webmaster, Direct, Wordstat, the site through MCP, read-only access to the OMS. Pydantic AI + aiogram.
Product photos from the factory and clients' offices are reworked with Gemini into clean catalog shots; product descriptions are generated from raw parameters.
B2B catalog for made-to-order furniture: configure a product in the calculator, see the price, order without registering.
Several mailboxes into Telegram, LLM triage, reply straight from the chat.
Booking of display showcases for rent.
Personal AI assistant: server, mail, Telegram, an Obsidian knowledge base.
Mail, ELK alerts; many one-task bots that were switched off once the task was done.
Arduino, 3D printing, home brewing with a temperature probe in the tank logging to a console, a hydroponics attempt, lamps with touch buttons — hardware has been a long-running interest, not a course.
I work AI-first.
Features are built with Claude Code; I review everything and write the critical parts myself — auth, payments, infrastructure — and spend most of my time on architecture and the product. Minimal dependencies, pragmatic simplicity over over-engineering, open standards, resilience to network failures. The clearest proof: the order management system we once built with a hired developer, I rewrote alone in a week.
Wrote specs and ran a development project (OMS v1); worked with freelance designers on the Pioner site — set tasks and signed off on layouts [TBD: details].