Core Principles

1. Systems Thinking

I work comfortably across all layers of the stack:

┌───────────────────────────────────────────┐
│  Application (Go, Python, Vue.js)         │
│  - REST APIs, CLI clients, Web UI         │
├───────────────────────────────────────────┤
│  System (Linux, Docker, systemd)          │
│  - Daemons, IPC, network tunneling        │
├───────────────────────────────────────────┤
│  Kernel (Netfilter, TUN/TAP, POSIX)       │
│  - Kernel modules, sockets, syscalls      │
└───────────────────────────────────────────┘

Example: VPN Manager project required all three layers:

  • Kernel: TUN/TAP interface, Netfilter hooks
  • System: vpnmd daemon (root privileges), systemd service
  • Application: vpnm CLI (Python), anyd IPC framework

2. Architecture First, Code Second

Before writing code, I define:

Question Example (MHA)
Problem Detect CIR changes across cloud providers
Constraints Different APIs, vendor-specific fields
Trade-offs Hash collisions vs. comparison speed
Success Metrics 100% change detection, <100ms hash calculation

Result: 5 documented specifications before implementation.

3. Verification Over Assumption

Layer Tool Purpose
Input Universe Validation Stack Validate Unified Models
Output Prometheus, Grafana Monitor system behavior
Behavior Notion specs, Habr articles Document decisions

4. Teach to Learn

I solidify understanding by teaching:

Platform Output Impact
MIPT 11 Go lectures 100+ students
JetBrains Academy Python Core project (Multilingual Translator) Educational resource for Hyperskill track
Habr 9 articles 237K views, 13K karma
Open Source 6 projects 1,756 GitHub contributions

5. Iterate Publicly

Share Early → Accept Feedback → Improve → Repeat
↓              ↓              ↓
GitHub        Issues, PRs    Real-world usage
Habr          Comments       237K views
PyPI          Downloads      anyd 0.4.1

← Back to About