I am a System Backend Engineer dedicated to building high-performance distributed systems and automation tools. I have a deep curiosity about system internals—ranging from OS concepts and networking protocols to Inference Runtime optimization.
- Multi-stack Backend Development: Experienced in Go, Python, Java, and PHP, designing scalable microservices and robust enterprise systems.
- AI Engineering: Exploring Inference Runtime (e.g., vLLM, TensorRT, Triton) for production-grade model deployment and performance tuning.
- Web Automation: Creator of cdpkit, a library leveraging Chrome DevTools Protocol for complex automation and large-scale data pipelines.
"Understanding the system is the foundation of writing better software."
- System Backend Engineering:
- Go: Distributed architecture, high-concurrency patterns, system utilities.
- Python: AI integration, high-performance APIs (FastAPI), data processing.
- Java / PHP: Scalable enterprise services and business logic implementation.
- AI & Inference:
- Deep dive into vLLM, Triton Inference Server, TensorRT, and Torch-TensorRT.
- Building automated STT (Speech-to-Text) processing pipelines.
- Infrastructure & DevOps: Docker, Kubernetes (GKE), Nginx, Terraform (IaC), and Zabbix monitoring.
- Low-level Exploration: C++ system tools, Linux kernel concepts, and algorithmic optimization.
- 🧩 cdpkit – Go-based CDP abstraction layer for efficient browser automation.
- 🎙️ stt-ai – High-performance STT pipeline integrated with Go, TS, and Redis.
- ⚡ Inference Lab – Benchmarking and optimizing vLLM, TensorRT, and Triton for production workloads.
- 🏗️ Distributed Scheduler – Designing reliable task scheduling systems and microservice orchestration.
- 🧠 Algorithms – Refining algorithmic intuition and performance optimization in C++.
- ✉️ Email: peggydevork@gmail.com
- Feel free to reach out via GitHub issues or PRs.