MR
mubdiur@portfolio:~$

// I build systems that work at scale — from automation pipelines to AI-powered tools.

0+ yearsCoding Experience
0+Projects Delivered
0.0%Uptime Maintained
0/7Production Systems
capabilities

Tech Stack

$ ls -la /usr/local/stack

Languages

TypeScriptPythonRustC#JavaKotlin

AI / ML

LLMsRAGLangChainOpenCVQdrantPyTorch

Automation

PlaywrightCI/CDWebSocketBashGitHub Actionssystemd

Infrastructure

Docker SwarmTraefikPostgreSQLRedisQdrantNginx
portfolio

Notable Work

$ cat /var/log/projects.log | tail -6

Portal Monitor

Automated extraction of structured time-series data from monitoring dashboards — replacing human monitoring engineers by scraping Dynatrace, ELK/Kibana, and Grafana UIs.

PlaywrightTypeScriptDockerNode.jsData Extraction

Site Monitor

Multi-agent production monitoring platform with health checks, circuit breakers, deduplicated alerting, and self-healing across distributed services.

Node.jsWebSocketPostgreSQLDockerAlerting

Knowledge Retrieval

Agentic RAG system with multi-hop reasoning across vector search (Qdrant), BM25 keyword search, and PostgreSQL for structured queries.

PythonLangChainQdrantLLMsFastAPI

PerfCat

Cross-platform system performance monitor tracking CPU, memory, disk, and network with configurable thresholds and Telegram-based alert delivery.

RustTelegram APIsystemdCLIMetrics

WinAlert

Native Windows desktop application for incident management with severity-based visual and audio notifications, triage workflow, and system tray integration.

C#.NET 8WPFTCPMVVM

Snipshot

Real-time screen rectangle detection overlay using computer vision — highlights detected rectangular regions under mouse cursor via transparent overlay.

Node.jsOpenCVComputer VisionBunDesktop
connect

Get in Touch

$ ping mubdiur@dev — 64 bytes from ambition: ttl=42

$ ./send_message.sh

Drop your email and I'll get back to you.