I build practical software systems around backend engineering, distributed architecture, real-time applications, and infrastructure.
I'm a full-stack software engineer with a strong focus on backend systems, infrastructure, and scalable application architecture.
I spend most of my time building systems with Node.js, designing APIs and services, working with databases and message brokers, integrating real-time communication, and deploying applications through containerized infrastructure.
I prefer solutions that are:
- Practical over unnecessarily complex
- Observable instead of impossible to debug
- Modular without blindly splitting everything into microservices
- Maintainable beyond the first successful deployment
- Designed for production, not just for demos
Backend Systems
├── REST APIs & service architecture
├── Authentication & authorization
├── Background jobs & queues
├── Real-time events
└── External system integrations
Distributed Applications
├── Microservices
├── Event-driven architecture
├── Kafka / RabbitMQ
├── Redis Pub/Sub
└── Service coordination
Infrastructure
├── Dockerized deployments
├── Linux servers
├── Reverse proxies
├── PostgreSQL / MongoDB
└── Production troubleshooting
Real-Time Communication
├── WebSocket
├── Server-Sent Events
├── SIP
├── Asterisk PBX
└── Browser-based softphones
I'm especially interested in systems where software meets infrastructure.
That includes:
- Distributed and event-driven systems
- AI-assisted applications and agent infrastructure
- Real-time communication systems
- Developer tooling
- Simulation and engineering software
- Observability and production reliability
- Infrastructure automation
- High-scale backend architecture
I also enjoy exploring technologies outside my usual stack when the problem is interesting enough.
backend:
- Node.js
- distributed systems
- API architecture
- asynchronous workloads
frontend:
- React
- Vue 3
- modern web applications
infrastructure:
- Docker
- Linux
- PostgreSQL
- Redis
- Nginx
real_time:
- WebSocket
- SSE
- SIP
- Asterisk
exploring:
- AI agents
- long-term AI memory
- engineering simulation
- developer infrastructure| Primary Editor | Visual Studio Code |
| C# Development | Visual Studio |
| Daily OS | Windows |
| Server Environment | Linux |
| Deployment | Docker |
| Primary Runtime | Node.js |
Lossless long-term memory for AI agents with verbatim conversation storage, hybrid retrieval, and MCP-native access.
Aerospace digital engineering workbench for scenario-driven SIL simulation, 6-DOF dynamics, flight software validation, and deterministic replay.


