MD FAHEEM AKRAM

Passionate about the convergence of cloud-native applications and AI, driven by the belief that this technology intersection will fundamentally reshape how we solve global challenges. I'm energized by building elegant systems that democratize powerful tools, making cutting-edge capabilities accessible to developers and communities worldwide.

1st State Rank, NEE I Examination (2018), Achieved the top state rank from Arunachal Pradesh in the NERIST Entrance Examination (NEE I) in 2018, demonstrating exceptional proficiency in Physics, Chemistry, and Mathematics.

Built an enterprise-grade Golang tool leveraging Kubernetes client-go to ingest Volcano Job specs and orchestrate ML training across multiple Kubernetes clusters — essentially a scheduling super-brain that auto-detects PyTorch, TensorFlow, and Spark workloads, profiles resource demands, and computes optimal GPU/CPU placement for maximum performance and serious cost savings. Designed intelligent placement algorithms (with Lua scripting hooks) that produce automated optimization strategies with proven 20–60% efficiency gains, engineered a comprehensive validation framework and battle-tested CLI, and basically turned an obsession with distributed systems, Golang craftsmanship, and cluster scheduling into a production-ready solution that makes large-scale training jobs run blazingly fast and far cheaper.

Made a full 2D space shooter game in Go using the Ebiten game engine — ships, lasers, explosions, the works. I built smart collision detection with spatial partitioning so it stays fast even when there’s a ton happening on screen. Created a scene management system that handles smooth transitions between menus, gameplay, and pause screens while keeping the game state intact. Added audio, sprite animations, and cross-platform asset handling so it feels polished on any device. Followed core game dev principles like game loops, physics simulation, and responsive controls, and even used Go’s modern features like Go embed to bundle assets right into the build for easy cross-platform play.

Built a federated multi-cluster Kubernetes application using Karmada to orchestrate Go microservices across 3 geographically distributed clusters, implementing advanced propagation policies and cross-cluster service discovery to explore cloud-native distributed systems architecture. Developed deep hands-on experience with Kubernetes primitives while diving into Karmada's resource interpretation and scheduling mechanisms, sparking my passion for multi-cluster orchestration and desire to contribute to CNCF projects. Currently expanding my knowledge of Karmada's internal architecture and resource interpreters, particularly interested in how Volcano scheduler integration enhances workload distribution in federated environments.