I build and ship production agentic systems. Some of what I've built:
- aeon-main — a multi-agent runtime I architected from scratch (~26k LOC, 270+ tests): persistent agents, background task scheduling, and a custom MCP server that lets agents spawn sub-agents recursively, past Claude Code's native 2-layer limit. Public.
- InBetween (CTO & co-founder) — an MCP server + tooling that lets AI agents message each other directly. In production.
- DOT — a full agentic framework I built from scratch that runs an autonomous Twitter account end to end. Public on GitHub.
- Claudeblox — a system of AI agents that build and playtest Roblox games on their own.
- Telegram integrations at the MTProto protocol level (Telethon), not just the Bot API.
Open to agentic-infrastructure contracts or a founding-engineer role.
Remote: Yes
Willing to relocate: No
Technologies: Python, TypeScript, Claude Code, custom MCP servers, multi-agent orchestration, RAG, Telegram MTProto/Bot API, Railway, Docker, Supabase
Résumé: https://sand0vvv.notion.site/Vladimir-Shiryaev-36c032b359888...
GitHub: https://github.com/sand0vvv
Email: vova.vova.shir@gmail.com
I build and ship production agentic systems. Some of what I've built:
- aeon-main — a multi-agent runtime I architected from scratch (~26k LOC, 270+ tests): persistent agents, background task scheduling, and a custom MCP server that lets agents spawn sub-agents recursively, past Claude Code's native 2-layer limit. Public.
- InBetween (CTO & co-founder) — an MCP server + tooling that lets AI agents message each other directly. In production.
- DOT — a full agentic framework I built from scratch that runs an autonomous Twitter account end to end. Public on GitHub.
- Claudeblox — a system of AI agents that build and playtest Roblox games on their own.
- Telegram integrations at the MTProto protocol level (Telethon), not just the Bot API.
Open to agentic-infrastructure contracts or a founding-engineer role.
reply