Software Engineer / Full-Stack Engineer

I'm Musab Bin Zafar

I build reliable backend services, full-stack applications, Python apps, and AI solutions with .NET, Java, Kotlin, React, and Next.js.

Musab Bin Zafar portrait
C#
.NET
Java
Kotlin
Spring Boot
React
Next.js
TypeScript
Python
PostgreSQL
SQL Server
MongoDB
Redis
Docker
Kubernetes
AWS
C#
.NET
Java
Kotlin
Spring Boot
React
Next.js
TypeScript
Python
PostgreSQL
SQL Server
MongoDB
Redis
Docker
Kubernetes
AWS

Recent Works

Software Engineer

I started my career in UX, product management, and App Store Optimization, where I learned how people discover and use products, and how to connect user behavior with business goals. That product mindset still shapes how I engineer.

After a two-year career break, I returned to technology through software engineering, working in Java, Scala, Kotlin, and Python. Reactive programming drew me in because it builds responsiveness and resilience into a system’s design.

Today I build high-volume .NET APIs and data-intensive applications using C#, Entity Framework, and SQL Server, alongside Java and Scala services on Spring Boot and Akka. I also create full-stack experiences with JavaScript and React, focusing on simple user journeys backed by maintainable architecture, thoughtful API design, and automated testing.

With a background in machine learning and deep learning, I’ve built microservices-based ML systems in Python and now work with generative and agentic AI, applying it to practical workflows while keeping product goals and system design aligned.

I ship and operate software with Git, Docker, Kubernetes, AWS, and CI/CD through GitHub Actions. I value collaboration, mentoring, and continuous learning, and I want to build products that solve meaningful problems at scale.

Current focus

Backend Engineering

1
  • C# / .NET
  • Java / Spring Boot
  • Kotlin / Spring WebFlux
  • REST APIs
  • GraphQL APIs

Blog

Writing

Article

From Vector RAG to Materialized State: Rethinking How Agents Remember

Explores why semantic retrieval is a poor fit for conversational memory, proposing event-driven compaction into structured state and a cache-aware context layout.

Read blog

Article

Building Large AI Systems Is an Architecture Problem, Not a Model Problem

Outlines architecture practices for maintainable production AI, including bounded contexts, clean boundaries, event-driven workflows, and continuous evaluation.

Read blog

Certifications

Courses and certificates.

Udemy

Spring WebFlux Masterclass: Reactive Microservices

Udemy

Reactive Applications with Spring WebFlux Framework

LinkedIn Learning

Kotlin Essential Training: Object-Oriented and Async Code