Skip to content
View maishivamhoo123's full-sized avatar

Organizations

@Project-HAMi

Block or report maishivamhoo123

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
maishivamhoo123/README.md

πŸ‘‹ Hi, I'm Shivam Kumar

Software Developer | HAMi (CNCF) Community Member | Go Enthusiast

I'm a backend-focused software developer driven by problem solving and systems reliability. I specialize in building robust backend systems in Go and contributing to world-class open-source infrastructure β€” with 50+ merged PRs across projects like Google, HAMi (CNCF), and the wider Kubernetes ecosystem. I currently serve as a Community Member on HAMi's Community Leadership team.


πŸ’Ό Experience

** Community Member β€” HAMi (CNCF)** Remote Β· February 2026 – Present

  • Drive feature development and code reviews as part of HAMi's Community Leadership team, helping steer technical discussions and community initiatives.
  • Optimized distributed node-lock scalability with exponential backoff and resolved a Prometheus metrics cardinality explosion to keep AI workloads highly available.
  • Hardened Kubernetes scheduler API routes against DoS and fixed kernel-level edge cases in NVIDIA hardware health checks.
  • Built a full GPU scheduling test environment (KIND cluster + nvml-mock on WSL2/Linux) that needs no physical GPU hardware; represented HAMi at the KubeCon India booth.

Open Source Contributor β€” Google (go-github), Apache, Kubernetes ecosystem Remote Β· March 2024 – Present

  • 50+ merged pull requests resolving backend and infrastructure issues across high-impact repositories.

Software Developer Intern β€” Jijiwisha Society Kanpur, UP Β· June 2024 – August 2024

  • Built a full-stack MERN website with stakeholders, growing NGO visibility and engagement by 25%.

πŸ› οΈ Proof of Work (Open Source Contributions)

πŸ›‘οΈ Security & System Reliability

  • Project-HAMi (#1620): Implemented DoS protection by introducing io.LimitReader to scheduler routes, preventing resource exhaustion from oversized payloads.
  • Project-HAMi (#1663): Optimized nodelock scalability with exponential backoff and listers, improving high availability for AI workloads under heavy concurrent load.
  • HAMi-core (#168): Added retry logic for nvmlInit so a transient GPU init failure no longer strands a process in a permanent failure state.
  • Project-HAMi (#1810): Fixed Kernel 6.17 handshake edge cases in NVIDIA hardware health checks.
  • HAMi-DRA (#25): Added OwnerReferences to ResourceClaims so orphaned claims are correctly garbage-collected.
  • Gogs (#8091): Engineered safety checks to ensure data integrity by verifying directory existence before repository initialization.
  • Syncthing (#10541): Enhanced system transparency by ensuring full device ID logging during discovery server startup.

πŸš€ Core Logic & Feature Engineering

  • Google (go-github #3891): Implemented omitzero support for BypassActors, leveraging new Go 1.24 features for more efficient API data handling.
  • Google (go-github #3944): Added support for organization artifact metadata APIs, including deployment/storage record endpoints with full test coverage.
  • Google (go-github #4288): Added enterprise billing usage endpoints and response types.
  • Google (go-github #4241): Added support for getting Copilot cloud agent configuration.
  • Apache (answer #1464): Added support for .env configurations to allow for dynamic, environment-specific application setups.
  • Gogs (#8089): Developed custom post-sign-out redirection logic to streamline user authentication workflows.
  • Syncthing (#10527): Standardized version output logic in the discovery server to improve CLI consistency.

πŸ› Optimization & User Experience

  • Project-HAMi (#1628): Fixed Prometheus metric cardinality explosion by optimizing container memory descriptor labels, reducing excessive time-series generation and improving monitoring stability.
  • ChainSafe (lodestar #8298): Optimized Ethereum consensus client logging for higher performance and cleaner telemetry data.
  • Termix-SSH (#182): Integrated dynamic versioning for CLI menus using environment variables.
  • OHC Network (#12197): Resolved critical responsiveness issues in healthcare clinical modules.
  • Memos (#5323): Enhanced UI accessibility and dark-mode contrast for improved user experience.
  • KAI-Scheduler (#1854): Fixed CI upgrade-version resolution by checking OCI chart availability.

πŸ“š Documentation & Community

  • Project-HAMi (#2064): Authored the design document for init-container GPU resource accounting.
  • HAMi website (#610): Added the Lab 10 GPU topology-aware scheduling tutorial, resolving a version-skew issue found along the way.
  • Reviewed 5+ pull requests as part of HAMi's feature working group and regularly attend community meetings.

Projects

βš–οΈ DSA-Powered HTTP Load Balancer (Go)

Dec 2025

A high-performance Layer 7 HTTP Load Balancer built in Go. Implements a Min-Heap (Priority Queue) to efficiently route traffic to the least-loaded backend server.

πŸš€ Features

  • Least Connections algorithm using Min-Heap
  • O(1) best-server lookup (heap root)
  • O(log N) rebalancing on connection updates
  • Concurrency-safe with sync.RWMutex
  • Active health checking (self-healing nodes)
  • Reverse proxy using net/http/httputil
  • Config-driven backend registration
  • Real-time status dashboard

🧠 Core Idea

Instead of scanning all servers (O(N)), the load balancer maintains a Min-Heap where priority = active connections.

  • Peek root β†’ Best server (O(1))
  • Increment/Decrement connections β†’ heap.Fix() (O(log N))

This ensures scalable traffic distribution as backend pool grows.

πŸ€– Microservices AIOps Platform

Feb 2026

An end-to-end AIOps platform built with microservices architecture. Detects anomalies, predicts incidents, and performs automated root cause analysis (RCA) in real time.

πŸ— Architecture

Services:

  • API Gateway (Go)
  • Auth Service (Go)
  • Worker Service (Go)
  • Anomaly Detector (Python - Isolation Forest)
  • Incident Predictor (XGBoost)
  • RCA Service
  • Prometheus + Grafana + Loki stack

All services run via Docker Compose.

✨ Features

  • Real-time anomaly detection (Isolation Forest)
  • Incident prediction using XGBoost
  • Automated Root Cause Analysis (log + metric correlation)
  • Prometheus metrics scraping
  • Centralized logging via Loki
  • Grafana dashboards
  • Synthetic traffic & chaos simulation

πŸ† Achievements

  • πŸ₯ˆ Runner-up, ABB Accelerator Hackathon 2026 β€” ranked 2nd among 14,000+ participants nationwide.
  • πŸ₯‡ Winner, Open Source Hackathon 2025 β€” among 500+ participants.
  • πŸ’» 650+ DSA problems solved; ranked 379th in LeetCode Weekly Contests (top 0.9%) with a peak rating of 1800+.
  • 🌐 Website developer for TEDx NIT Andhra Pradesh (2024, 2025), Techkriya (2024), and TechFest (2024).

πŸŽ“ Education

National Institute of Technology, Andhra Pradesh B.Tech, Mechanical Engineering Β· June 2023 – April 2027 Relevant coursework: Data Structures & Algorithms, Object-Oriented Programming, Artificial Intelligence, Distributed Systems.


🧰 Tech Stack

  • Languages: Go (Golang), Java, Python, JavaScript, SQL
  • Frameworks & Other: React, CUDA
  • Cloud & DevOps: Docker, Kubernetes, AWS (EC2, IAM, S3, VPC, CloudWatch), Git, GitHub, Linux/Bash, WSL

🀝 Connect with me


I am currently seeking a Software Engineering Internship (Backend/DevOps/Go). If you like my work, let's talk!

Pinned Loading

  1. Project-HAMi/HAMi Project-HAMi/HAMi Public

    Heterogeneous GPU Sharing on Kubernetes

    Go 4.2k 716

  2. google/go-github google/go-github Public

    Go library for accessing the GitHub v3 API

    Go 11.3k 2.5k