Python engineer. I build async backends, real-time integrations and automation — and lately the tooling that lets AI models actually do things: MCP servers, agent memory, multi-agent coordination.
I like software that is boring in the right places: typed, tested, and predictable under load.
class Rufus011:
focus = ["async backends", "real-time systems", "AI agents & MCP"]
stack = ["Python", "asyncio", "socket.io", "pydantic", "pytest"]
tooling = ["ruff", "mypy", "GitHub Actions", "Docker"]
principle = "typed · tested · no secrets in the repo"
speaks = ["ru_RU", "en_US"]
def looking_for(self):
return "interesting automation problems · open-source collaboration"%%{init: {'theme':'dark','themeVariables':{'primaryColor':'#0d1117','primaryTextColor':'#c9d1d9','primaryBorderColor':'#00ff88','lineColor':'#00ffcc','fontFamily':'monospace'}}}%%
flowchart LR
R(("RUFUS_011"))
A["⚙️ Async backends<br/>asyncio · websockets"]
B["🤖 AI agents<br/>MCP servers · tooling"]
C["🔌 Integrations<br/>REST · socket.io · scrapers"]
D["🧪 Quality<br/>pytest · mypy · ruff · CI"]
R --> A & B & C & D
style R fill:#0d1117,stroke:#00ff88,color:#00ff88
style A fill:#11161d,stroke:#00ff88,color:#c9d1d9
style B fill:#11161d,stroke:#00ffff,color:#c9d1d9
style C fill:#11161d,stroke:#48b0ff,color:#c9d1d9
style D fill:#11161d,stroke:#00ff88,color:#c9d1d9
Async backends — event-driven clients that stay alive: reconnects, backpressure, no polling loops. AI agents — MCP servers, persistent memory, and coordination so several models can work together. Integrations — talking to APIs that don't want to be talked to. Quality — offline test suites, strict typing, CI that actually catches things.
Open to collaboration and interesting problems. Everything I publish is MIT and free — a ⭐ on a repo is already plenty.
